1930 Gift of Richard H. Backus May, 1988 V' /2M ^.P.Jd^ DEPARTMENT OF COMMERCE BULLETIN OF THE ^L ^^A UNITED STATES BUREAU OF FISHERIES o= Xs I "^ I -0 '. 3- : 3- i □ = CD = r^ - m = a VOL. XL 1924 IN TWO PARTS— PART I HENRY O'MALLEY COMMISSIONER PRICE, $2.00 (Buckram) Sold only by the Superintendent of Documents, Government Printing Office, Washington, D. C. WASHINGTON GOVEFINMENT PREMTING OFFICE 1925 O ll ^ Bureau of Fisheries Document No. 965 FISHES OF THE GULF OF MAINE By HENRY B. BIGELOW, Museum oj Comparative Zoology, Harvard University, and WILLIAM W. WELSH, Late Scientific Assistant, U. S. Bureau of Fisheries CONTENTS Page Introduction 6 Scope of the work 1 7 Area covered 7 Sources of information 8 Use of the keys 10 Key to Gulf of Maine fishes 11 The lampreys. Class Marsipobranchii 15 The hagfishes and lampreys. Familes Myxinidae and Petromyzonidse 3 5 Hagfish {Myxine glutinosa Linnaeus) 16 Sea lamprej' {Petromyzon marinus Linnaeus) 18 True fishes. Class Pisces 21 Sharks and rays. Subclass Elasmobranchii 21 Sharks 22 The eel sharks. Family Chlamydoselachidae 24 Eel shark (Chlamydoselachus anguineus Garman) 24 The smooth dogfishes. Family Galeorhinidae 25 Smooth dogfish (Galeorhinus tewis Valmont) 25 Requiem sharks. Family Carcharinidae 27 Tiger shark (Galeocerdo arclicus Faber) 27 Great blue shark (Galeus glaucus Linnseus) 28 Dusky shark {Carcharhinus obscurus LeSueur) 29 The hammer-headed sharks. Family Cestraciontidse 30 Hammerhead shark {Cestracion zygxna Linnaeus) 31 The thresher sharks. Family Vulpeculidfe 32 Thresher {Vulpecula marina Valmont) 32 The sand sharks. Family Carchariidae 34 Sand shark {Carcharias taurus Ksi&nesque) 34 The mackerel sharks. Family Isuridae 35 Mackerel shark {Isurus punctalus Storer) 36 Sharp-nosed mackerel shark {Isurus tigris Atwood) 38 White shark {Carcharodon carcharias Linnaeus) 39 Basking shark {Cetorhinus maximus Gunner) 41 The spiny dogfishes. Family Squalidae 44 Spiny dogfish {Squalus acanthias Linnjeus) 44 Portuguese shark {Centroscymnus coelolepis Bocage and Capello) 51 Black dogfish {Centroscyllium fabricii Reinhardt) 52 1 1 V*0(Kb HOa. MASS. \ 2 BULLETIN OF THE BUREAU OF FISHERIES True fishes — Continued. Sharks and rays — Continued. Sharks — Continued. Pago The nurse sharks. Family Scymnorhinidoe 53 Greenland shark (Soinniosus microcephalus Bloch and Schneider) 53 The bramble sharks. Family Echinorhinidae 55 Bramble shark (Echinorhinus brucus Bonnaterre) 55 Skates and rays 56 The skates. Family Rajidse 58 Little skate {Raja erinacea Mitchill) 58 Big skate {Raja diaphanes Mitchill) 60 Prickly skate {Raja scabrata Garman) 62 Brier skate {Raja eglanteria Bosc) 64 Smooth skate {Raja senla Garman) 65 Barn-door skate {Raja slabuliforis Garman) 66 The torpedoes. Family Narcaciontidffi 68 Torpedo {Narcacion nobilianus Bonaparte) 68 The sting rays. Families Dasybatidse and Myliobatidae 70 Sting ray {Dasybatus niarinus Klein) 70 Sting ray {Dasybatus hastatus DeKay) 70 Cow-nosed ray {Rhinoptera quadriloba LeSueur) 72 Chimseroids. Subclass Holocephali 73 The chimaeras. Family Chimaeridee 73 Chimsera {Chimxra affinis Capello) 73 The bony fishes. Subclass Teleostomi 74 The sturgeons. Family Acipenseridse 74 Sturgeon {Acipenser sturio Linnaeus) 74 The eels. Families Anguillidaj, Synaphobranchidae, LeptocephaUdae, Simenchelyidae, and Nemichthyidae 77 Eel {Anguilla rostrata LeSueur) 78 Shme eel {Simenchelys parasiticus Gill) 83 Long-nosed eel {Synaphobranchus pinnalus Gronow) 84 Conger {Leptocephalus conger Linnaeus) 86 Snipe eel {Nemichthys scolopaceus Richardson) 88 The tarpons and herrings. Families Elopidoe and Clupeidae 90 Tarpon {Tarpon atlaniicus Cuvier and Valenciennes) 91 Round herring {Etrumeus teres DeKay) 91 Herring {Clupea harengus Linnaeus) "2 Hickory shad {Pomolob us mediocris Mitchill) 105 Alewife {Pomolobus pseudoharengus Wilson) 107 Blueback {Pomolobus xstivalis MitchiU) HO Shad {Alosa sapidissima Wilson) 113 Menhaden {Brevoortia lyrannus Latrobe) 118 ■ The anchovies. Family Engraulididae 124 Anchovy {Anchovia mitchilli Cuvier and Valenciennes) 124 The salmons. Family Salmonidae 126 Humpback salmon {Oncorhynchus gorbuscha Walbaum) 126 Salmon {Salnio salar Linnaeus) 130 Brook trout {Salvelinus fontinalis Mitchill) 138 The smelts. Family Argentinidae ^ 140 Capelin {Mallotus villosus Miiller) 140 Smelt {Osmerus mordax Mitchill) 143 Argentine {Argentina situs Ascanius) 14' --I FISHES OF THE GULF OF MAINE 3 True fishes — Continued. The bony fishes — Continued. Page The lanternfishes. Family Myctophidse 149 Lanternfish (Mthoiprora effulgens Goode and Bean) 149 Family Stomiatidse 151 Pearlsides {Maurolicus pennanli Walbaum) 151 Viperfish (Chauliodus sloanei Bloch and Schneider) 153 Cj'clothone {Cydothone signata Garman) 153 The lancetfishes. Family Alepisauridae 154 Lancetfish (Alepisaurus ferox Gill) 154 The mummichogs or killifishes. Family Poecilidae 155 Common mummichog {Fundulus heteroclitus Linnaeus) 156 Striped mummichog (Fundulus majalis Walbaum) 158 Sheepshead minnow (Cyprinodon variegaius lia,c6pkde) 159 The billfishes. Family Belonidaj 161 Silver gar {Tylosurus marinus Walbaum) 161 The halfbeaks. Family Hemiramphidje 163 Half beak (Hyporhamphus roberti Cuvier and Valenciennes) 163 The needlefishes. Family Scomberesocidse 164 Needlefish {Scomhercsox saurus Walbaum) 164 The sticklebacks. Family Gasterosteidie 166 Nine-spined stickleback (Pungitius pungitius Linnteus) 167 Three-spined stickleback (Gaslerosteus aculealus Linnaeus) 168 Two-spined stickleback (Gaslerosteus bispinosus Walbaum) 171 Four-spined stickleback (Apeltes quadracus MitcMU) 171 The trumpetfishes. Family Fistulariidre 173 Trumpetfish (Fislularia labacaria Linnaeus) 173 The pipefishes. Family Syngnathidae 174 Pipefish (Siphostoma fuscum Storer) 175 The sea horses. Family Hippocampidae 177 Sea horse (Hippocampus hudsonius DeKay) 177 The silversides. Family Atherinidae 178 Silverside (Menidia nolala Mitchill) 179 Waxen silverside (Menidia beryllina cerea Kendall) 181 The mullets. Family Mugilidae 182 Mullet (Mugil cephalus, Linnaeus) 182 The sand launces. Family Ammodytidae 183 Sand launce (Ammodytes americanus DeKay) 183 The mackerels. Family Scombridce 187 Mackerel (Scomber scomhrus Linnaeus) 188 Chub mackerel (Pneumalophorus colias Gmelin) 209 Striped bonito (Gymnosarda pelamis Linnaeus) 211 Tuna (Thunnus Ihynnus Linnaeus) 212 Common bonito (Sarda sarda Bloch) 215 Spanish mackerel (Scomberomorus maculatus Mitchill) 217 King mackerel (Scomberomorus regalis Bloch) 219 The escolars. Family Gempylidae 220 Escolar (Ruvellus pretiosus Cocco) 220 The cutlasfishes. Family Trichiuridae 220 Cutlasfish (TrichiuTus leplurus Linnaeus) 220 The swordfishes. Family Xiphiidae 221 Swordfish (Xiphias gladius Linnaeus) 221 4 BULLETIN OF THE BUREAU OF FISHERIES True fishes — Continued. The bony fishes — Continued. Page The sailfishes. Family Istiophoridae 227 Spearfish {Teirapiurus imperator Blooh and Schneider) 227 The pompanos. Family Carangidae 228 Pilotfish (Naucrates ductor Linnaeus) 229 Rudderfish {Seriola zonata Mitchill) 230 Mackerel scad (Decapterus macarellus Cuvier and Valenciennes) 232 Crevalle {Caranx hippos Linnaeus) 233 Hardtail (Caranx crysos Mitchill) 234 Moonfish {Vomer selapinnis Mitchill) 235 Lookdown (Selene vomer Linnaeus) 236 The bluefishes. Family Pomatomidae 237 Bluefish (Pomatomus sallatrix Linnaeus) 237 The mariposas. Family Lampridce 242 Opah (Lampris luna Gmelin) 242 The rudderfishes. Family Centrolophidoe 243 Barrelfish (Palinurichthys perciformis Mitchill) 243 The butterfishes. Family Stromateidae 245 Butterfish (Poronotus Iriacanthus Peck) 245 Harvestfish (Peprilus paru Linnaeus) 250 The sea basses. Family Serranidae 251 Striped bass (Roccus lineatus Bloch) 251 White perch (Morone americana Gmelin) 257 Sea bass (Centroprisles slriatus Linnaeus) 259 The Catalufas. Family Priacanthidae 261 Big-e3"e (Pseudopriacanthus altus Gill) 261 The sea breams or porgies. Family Sparidae 262 Scup (Stenotomus chrysops Linnaeus) 263 Sheepshead (Archosargus probatocephalus Walbaum) 268 The croakers or weakfishes. Family Sciaenidae 269 Weakfish (Cynoscion regalis Bloch and Schneider) 270 Kingfish (Menticirrhus saxalilis Bloch and Schneider) 277 Black drum (Pogonias cromis Linnaeus) 279 The cunners. Family Labridae 280 Gunner (Tautogolabrus adspersus Walbaum) 281 Tautog (Tauloga onilis Linnaeus) 286 The John Dories. Family Zeidae 291 John Dory (Zenopsis ocellatus Storer) 291 The triggerfishes. Family Balistidae 293 Triggerfish (Balisles carolinensis Gmelin) 293 The filefishes. Family Monacanthidae 294 Filefish (Monacanthus hispidus Linnaeus) 295 Filefish (Monacanthus ciliatus Mitchill) 296 Orange filefish (Alutera schcepfii Walbaum) 296 The puffers and porcupine-fishes. Families Tetraodontidae and Diodontidae. 297 Puffer (Spheroides maculatus Bloch and Schneider) 298 Burrfish (Chilomyderus schcepfii Walbaum) 300 The headfishes. Family Molidae 301 Sunfish (Mola mola Linnaeus) 301 The rockfishes. Family Scorpaenidae 304 Rosefish (Sebasies marinus hinnseus) 304 Black-bellied rosefish (Helicolenus maderensis Goode and Bean) 313 FISHES OF THE GULF OF MAINE 5 True fishes — Continued. The bony fishes — Continued. Page The sculpins. Family Cottidae 314 Hook-eared sculpin {Artediellus atlanlicus Jordan and Evermann) 314 Mailed sculpin (Triglops ommatistius GUbert) 316 Little sculpin (Myoxocephalus xneus Mitohill) 318 Shorthorn sculpin {Myoxocephalus scorpius Linnaeus) 320 Longhorn sculpin (Myoxocephalus octodecemspinosus Mitchill) 325 Staghorn sculpin {Gymnocanthus tricuspis Reinhardt) 328 Deep-sea sculpin {Cotlunculus microps Collett) 329 Sea raven (Hemilripterus americanus Gmelin) 330 The alligatorfishes. Family Agonidae 332 Alligatorfish (Aspidophoroides monopterygius Bloch) 333 The lumpfishes. Family Cyclopteridae 334 Lumpfish {Cyclopierus lumpus Linnaeus) 334 Spiny lumpfish (Eumicroiremus spinosus Miiller) 339 The sea snails. Family Liparidae 340 Sea snail {Neoliparis atlanticus Jordan and Evermann) 340 Striped sea snail (Liparis liparis Cu vier) 342 The sea robins or gurnards. Family Triglidae 344 Common sea robin (Prionoius carolinus Linnteus) 345 Red-winged sea robin (Prionoius slrigalus Cuvier and Valenciennes) 348 The remoras. Family Echeneididae 349 Shark sucker (Echeneis naucrates Linnaeus) 349 Swordfish sucker (Remora brachyptera Lowe) 350 Remora (Remora remora Linnaeus) 351 The tilefishes. Family Malacanthidae 352 Tilefish (Lopholalilus chamxleonticeps Goode and Bean) 352 The toadfishes. Family Batrachoididae 356 Toadfish (Opsanus tau Linnaeus) 357 The blennies. Family Blenniidae 359 Rock eel (Pholis gunnellus Linnaeus) 359 Snake blenny (Lumpenus lampelrspformis Walbaum) 363 Shanny (Leptoclinus maculalus Fries) 365 Radiated shanny ( Ulvaria subbifurcata Storer) • 366 The wrymouths. Family Cry ptacanthodidae 368 Wrymouth (Cryptacanlhodes maculalus Storer) 368 The wolffishes. Family Anarhichadidae 370 Wolffish (Anarhichas lupus Linnteus) 370 Spotted wolffish (Anarhichas minor Ol&fscn) 375 The eelpouts. Family Zoarcidae 376 Eelpout (Zoarces anguillaris Peck) 378 Wolf eel (Lycenchelys verrillii Goode and Bean) 382 Arctic eelpout (Lycodes reticulatus Reinhardt) 383 The cusk eels. Family Ophidiidae 384 Cusk eel (Lepophidium cervinum Goode and Bean) 384 The silver hakes and cods. Familes Merlucciidae and Gadidae, .- 385 Silver hake (Merluccius bilinearis Mitchill) 386 American pollock (Pollachius virens Linnaeus) 396 Tomcod (Microgadus tomcod Walbaum) 406 Cod (Gadus callarias Linnaeus) 409 Haddock (Melanogrammus ssglifinus Linnaeus) 432 Blue hake (Antimora viola Goode and Bean) 444 White hake (Urophycis tenuis Mitcliill) 446 6 BULLETIN OF THE BUREAU OF FISHERIES True fishes — Continued. The bony fishes — Continued. The silver hakes and cods — Continued. Page Squirrel hake {Urophycis c/i «ss Walbaum) 447 Spotted hake {Urophycis regius Walbaum) 455 Long-finned hake (Urophycis chesteri Goode and Bean) 456 Hakeling (Physiculus fulvus Bean) 457 Four-bearded rockhng {Enchelyopus cimbrius Linnaeus) 458 Cusk {Brosmius brosme Miiller) 462 The grenadiers. Family Macrouridse - 467 Common grenadier {Macrourus bairdii Goode and Bean) 468 Smooth-spined grenadier {Macrourus berglax Lac^pede) 470 Long-nosed grenadier {Ccelorhynchus carminatus Goode) 471 The flounders and soles. Famihes Pleuronectidae and Soleidae 472 Halibut {Hippoglossus hippoglossus Linnaeus) 473 Greenland hahbut {Reinhardtiiis hippoglossoides Walbaum) 481 American plaice {Hippoglossoides platessoides Fabricius) 482 Summer flounder {Paralichthys dentatus Linnaeus) 491 Four-spotted flounder {Paralichthys oblongus Mitchill) 494 Rusty dab {Limanda ferruginea Storer) 495 Winter flounder (Pseudopleuronectes americanus Walbaum) 501 Georges Bank flounder {Pseudopleuronectes dignabilis Kendall) 507 Smooth flounder {Liopselia putnaini Gill) 508 Witch flounder {Glyptocephalus cynoglossus Linnaeus) 511 Sand flounder {Lophopselta maculala Mitchill) 516 Gulf Stream flounder {Ciiharichthys arctifrons Goode) 521 Hogchoker {Achirus fasciatus Lac^pede) 522 The anglers. Family Lophiidae 524 Goosefish {Lophius piscatorius Linnaeus) 524 Bibliography 533 Addenda 551 Index 555 INTRODUCTION In the summer of 1912 the Bm-eau of Fisheries, with the cooperation of the Museum of Comparative Zoology of Harvard University, undertook an oceano- graphic and biological survey of the Gulf of Maine, with special reference to its fishes and floating plants and animals (plankton), its physical and chemical state, .and the circulation of its waters. Subsequent cruises were made on the fisheries schooner Grampus during the summers and autumns of 1913, 1914, 1915, and 1916, and during the winters and springs of 1913 and 1915. The work was interrupted by the war, but was resumed with a cruise of the fisheries steamer Albatross in the late winter and spring of 1920 and continued by the fisheries steamer Halcyon during the winter and spring of 1920-21 and the stmimers of 1921 and 1922. Several reports on special phases of the survey have been published, but not until 1920 did the body of data warrant undertaking a general account of the fish faiina, general biology, and oceanography of the Gulf, of which the present memoir is the first part. In the division of labor the preparation of the section on the fishes was assigned to my coworker, W. W. Welsh, who had given special attention to this phase of the work throughout all the years of the survey, both on the regular oceanographic cruises and on many trips on commercial fishing vessels, in the course of which he FISHES OF THE GULF OF MAINE 7 had gathered a large body of original observations on the growth, reproduction, diet, and other phases of the lives of many of the more important species. The report was far advanced when interrupted by his untimely death, and so much of the material had been collected that, at the request of the Bureau of Fisheries, I have undertaken to carry it to publication along the lines originally laid down, though I am unable to give it the value it would have possessed had Mr. Welsh been able to finish it. SCOPE OF THE WORK Our aim has been to prepare a handbook for the ready identification of the fishes occurring in the Gulf of Maine, and to present a concise statement of what is known of the distribution, relative abundance, and the more significant facts in the life history of each. The descriptions have been made as little technical as is compatible with scientific accuracy, and are chiefly limited to such external features as may suffice for identification in the field. As a further aid to identification, keys to all species have been provided. In every case the sizes of larval fish or eggs have been given in millimeters (1 inch equals 25.4 millimeters), but these can be easily converted into inches or parts of an inch. We have followed Garman (1913) in the nomenclature of the sharks, skates, and rays, and Jordan and Evermann (1896-1900) for all the others, except as noted. For each species we have given page references to these authors, where the reader, if interested, may find more detailed descriptions and synonymies. Most of the illustrations have been borrowed from earlier publications, but a few are original. Rules given under illustrations represent a length of 1 inch. AREA COVERED The term " Gulf of Maine " covers the oceanic bight from Nantucket and Cape Cod on the west to Cape Sable on the east, thus including the shore lines of northern Massachusetts, New Hampshire, Maine, and parts of New Brunswick and Nova Scotia. The eastern and western boundaries adopted in this paper are 65° and 70° west longitude, respectively. Southern species, recorded but once from Nantucket and which have no real status in the Gulf of Maine except as accidental stragglers, have been relegated to footnotes. The Gulf of Maine has a natural seaward rim formed by Nantucket Shoals, Georges Bank, and Browns Bank. We have chosen the 150-fathom contour as the arbitrary offshore boundary because this will include all the species likely to be caught by commercial fishermen and will exclude almost the entire category of deep-sea fishes so numerous in the basin of the open Atlantic but not constituents of the fauna of the Gulf of Maine. The general geography of this area wUl be the subject of another report, but it may not be amiss to point out here that the temperature of the Gulf and its fauna as a whole are boreal, its southern and western boundaries being the northern limit of common occuiTcnce of many southern species of fishes and invertebrates. 102274— 25t 2 8 BULLETIN OF THE BUREAU OF FISHERIES SOURCES OF INFORMATION The literature dealing with the fishes of tlie GuK of Maine begins with the earliest descriptions of New England, for the fishery possibilities of the Gulf so impressed the early voyagers, even prior to the first settlement, that almost all accounts of their travels contain first-hand observations on the local abundance of fish of one species or another. Capt. John Smith (1616), for instance, commented on the abundance of sturgeon, cod, hake, haddock, cole (the American poUock), cusk, sharks, mackerel, herring, cunners, eels, salmon, and bass in 1616, while Wood (1634), in his " New England's Prospect," gives much interesting information, some of which is quoted hereafter. It was not until the early part of the nineteenth century that the sea fishes of northern New England and of the Maritime Provinces began to attract scientific attention, but since then the local faunal lists for that region have become numerous. The following, in chronological order, are the most important of these: 1850. — "Report on the sea and river fisheries of New Brunswick, within the Gulf of St. Law- rence and Bay of Chaleur," by M. H. Perley. 137 pp., 1850. Fredericton. 1853-1867. — "A history of the fishes of Massachusetts," by David Humphreys Storer. Mem- oirs, American Academy of Arts and Sciences, New Series, Vol. V, pp. 49-92, 122-168, and 257-296; Vol. VI, pp. 309-372; Vol. VIII, pp. 389-439; Vol. IX, pp. 217-256, 39 pis. (Also in book form with supplement.) Cambridge and Boston. 1879. — "A list of the fishes of Essex County, including those of Massachusetts Bay, according to the latest results of the work of the U. S. Fish Commission," by George Brown Goode and Tarle- ton H. Bean. Bulletin, Essex Institute, Vol. XI, No. 1, pp. 1-38. Salem. 1884. — " Natural history of useful aquatic animals," by George Brown Goode and associates. Section I, The Fisheries and Fishery Industries of the United States, published jointly by the United States Fish Commission and the United States Bureau of the Census. Washington. 1908. — " Fauna of New England. 8. List of the Pisces," by William C. Kendall. Occasional Papers, Boston Society of Natural History, Vol. VII, No. 8, April, 1908, pp. 1-152. Boston. 1914. — "An annotated catalogue of the fishes of Maine," by William C. Kendall. Proceed- ings, Portland Society of Natural History, Vol. Ill, 1914, Part 1, pp. 1-198. Portland. 1922. — "The fishes of the Bay of Fundy," by A. G. Huntsman. Contributions to Canadian Biology, 1921 (1922), No. 3, pp. 1-24. Ottawa. Either at first hand or by reference to the original sources these faunal lists contain all the published locality records of the rarer species, while the last two, with a paper by Gill (1905b) , give complete ichthyological bibliographies respectively for the coasts of Maine, New Brunswick and Nova Scotia, and Massachusetts. A similar list of the captures of deep-water forms along the outer part of the Conti- nental Shelf is contained in Goode and Bean's "Oceanic Ichthyology" (1S96). The most pertinent extralimital lists are Smith's (1898) and Sumner, Osburn, and Cole's (1913) lists of Woods Hole fishes for the waters immediately to the west, and Halket's (1913) check list of the fishes of Canada for those to the east and north of the Gulf of Maine. With these readily available we have not thought it worth while to burden the present paper with the authorities for localities except in the more interesting cases. To save constant repetition we state here that almost all of the information as to the Bay of Fundy given hereafter is drawn either from Huntsman's paper or from his unpublished notes. Much information as to local FISHES OF THE GULF OF MAINE 9 distribution and relative abundance has been gleaned from the fishery statistics published by the United States Bureau of Fisheries, the Dominion of Canada, and the Commonwealth of Massachusetts. The literature dealing with the lives and habits of fishes occurring in the Gulf of Maine is very extensive, for most of the important commercial species, and many of the others, are common to both sides of the North Atlantic and have come within the scope of the intensive studies carried out of late years by European zoologists in conjunction with the International Committee for the Exploration of the Sea, while considerable attention has been devoted to them by American ichthyologists, also (published for the most part by the United States Bureau of Fisheries) . The many scattered accounts of eggs and larvae of northern fishes have been collected by Ehrenbaum ^ in his general summary of their developmental stages, a compilation the utility of which can hardly be overrated. Among the other general European manuals, Day's "Fishes of Great Britain and Ireland" ^ and Smitt's "Scandinavian Fishes" ^ are especially helpful. We have also had access to a great amount of unpublished material in the files of the Bureau of Fisheries, especially instructive being the schedules turned in by observers who accompanied certain otter trawlers during 1913, and the observations of Vinal Edwards on the diet of fishes at Woods Hole. The superintendents of the New England hatcheries have supplied much valuable information, as noted in the appro- priate connections. Dr. A. G. Huntsman has, with great kindness, contributed his unpublished notes on the fishes of the Bay of Fundy and Gulf of St. Lawrence, allowing us to quote freely from them, while Prof. J. P. McMurrich has permitted the use of his unpublished plankton records. W. F. Clapp, formerly of the Museum of Comparative Zoology at Harvard University, has contributed many interesting notes gleaned during his experience as a fisherman before his entrance into the scientific field. Harry Piers, of the Provincial Museum of Halifax, has supplied interesting notes on the occurrence of the blue shark. We owe a debt of gratitude, also, to Dr. Samuel Garman, who has ever been ready with assistance, and to W. C. Adams, director of the division of fisheries and game of the State of Massachusetts. Finally, we wish to express our thanks to the many commercial fishermen M'ho have unfailingly met our inquiries in the most cordial way and who supplied Mr. Welsh with a vast amount of first-hand informa- tion on the habits, distribution, and abimdance of the conamercial fishes, which could be had from no other source. Without their help the preparation of this handbook would have been impossible. ' Eier and Larven von Fischen, by E. Ehrenbaum. Nordisches Plankton, Vol. I, 1905-1909 (1911), 413 pp., 148 figs. EleJ und Leipzig. (Appeared in two parts as Lief. 4, 1905, and Lief. 10, 1909.) ' The fishes of Great Britain and Ireland, by F. Day. Text and atlas, 1880. London and Edinborough. ' A history of Scandinavian fishes, by B. Fries, C. V. Ekstrom, and C. Sundervall. Second edition revised and completed by F. A. Smitt, 1892, 1,240 pp.. 53 pis. Stockholm. 10 BULLETIN or THE BUREAU OF FISHERIES USE OF THE KEYS The various fins and other structures mentioned in the keys are named in the accompanying outline of a haddock, and the simplest way to explain the use of the keys is to use that species as an example, running it down with the outline at hand for reference. 1 stioisai f iTi Jl 2Tii(lorsd fiu Baibel • ?'' 2nd. Anal fin tstAna-lfin. Ventral fin Fig. 1.— Diagram of a haddock, explainmg terms used in key Turning to section 1, Key A (p. 11), the fish in question evidently fits the second alternative, for it has bony jaws and pectoral fins and is not formed like an eel, which refers us to section 3. As our fish does not have a tubular snout this leads us to section 4, and this in turn to section 5, since neither jaw is greatly pro- longed. There being only one gill opening on each side we go from section 5 to section 7, and from section 7 to section 8, for there is a distinct tail. Since the fish has no sucking plate on the top of the head this leads to section 11, and this in turn to section 12 because the tail fin is nearly symmetrical in outline. Since the fish is not clothed in an armor of bony plates we are referred by the second alternative of section 12 to section 13. In our specimen the anal fin is clearly separated from the caudal, and section 13 therefore refers us to section 14. As all the fins are supported by rays we must go on to section 15, and from there to section IS as there are no fleshy flaps or tags on the sides of the head.'* Our fish does not lie fiat on one side (that is, it is a round, not a flat fish) and therefore it fits the second alternative under section 18, which refers it to section 19, and as there is more than one separate dorsal fin, this leads to Key F, page 13. As we have already determined, all the dorsal fins have soft rays, and since there are no spines in any of the fins (a fact easily determined by feeling them) , this sends us to the key to the cod and silver hake families (p. 385). Turning to the first section of the latter we find that the fish fits the first alternative (3 dorsal fins and 2 anals), which refers it to section 2, and here the black lateral line and the dark blotch on each shoulder name it a haddock. Any other Gulf of Maine species is to be named in the same way, starting with section 1, Key A, and following through the appropriate alternatives as they refer it from section to section. * There is a barbel on the chin but this is very different in appearance from the skin flaps around the 3aws characteristic of the few species that fall under the first alternative. FISHES OF THE GXJLF OF MAINE 11 KEY TO GULF OF MAINE FISHES Key A 1. Mouth soft with no bony jaws; form eel-like; no pectoral fins 2 Mouth has bony jaws; pectoral fins present if form is eel-like 3 2. Two separate fins on back; no barbels on snout Lamprey, p. 18 Only one fin on back; barbels on snout Hag, p. 16 3. Bones of head fused in a tubular snout, with mouth at tip Key B, p. 12 No tubular snout 4 4. One or both jaws greatly prolonged as a bony sword or bill Key C, p. 12 Neither jaw greatly prolonged 5 5. Five or more pairs of large gill openings (on sides of neck in sharks; on lower surface in skates) . 6 Only one external gill opening, large or small, on each side 7 6. General form cylindrical Sharks, key, p. 23 General form flat and diskhke, with long whiphke tail Skates and rays, key, p. 57 7. Body abruptly square-cut close behind the very high dorsal and anal fins. Sunfish, p. 301 Body with distinct tail 8 8. Large sucking plate or disk, either ou the top of the head or on the chest 9 No sucking disk or plate 11 9. Sucking plate on top of head Remora family key, p. 349 Sucking disk on chest 10 10. General form like a tadpole; anal fin originates about as far back as the tip of the pectoral Sea snail family key, p. 340 General form not Uke a tadpole, but high arched with longitudinal ridges; anal fin originates far behind tip of pectoral , Lumpfish family key, p. 334 11. Tail like a shark, i. e., with upper lobe much longer than lower Sturgeon, p. 75 Tail nearly symmetrical 12 12. Whole head and body clothed in continuous armor of bony plates.. Alligator fish, p. 333 If bony plates are present they do not form a continuous armor over head and body 13 13. No clear separation between anal and caudal fins, which together form one contin- uous fin (anal portion may be either long or short) Key D, p. 12 Anal and caudal fins separated by a deep notch if not by a space 14 14. There is a fleshy (adipose) fin with neither spines nor rays behind the rayed dorsal fin Key E, p. 13 A fleshy flap* in front of dorsal fin Tilefish, p. 352 All dorsal flns supported by rays or spines, which can be felt if not seen; without fleshy lobee or adipose fins either in front of or behind them 15 15. Head fringed with fleshy tags or flaps; much broader than body 16 Head not fringed with fleshy flaps 18 16. Lower jaw projects far beyond upper, exposing very large conical teeth even when mouth is closed; two long isolated spines on top of head in front of eyes. Goosefish, p. 524 Lower jaw does not project noticeably beyond upper; teeth small; no long isolated spines in front of eyes 17 17. First (spiny) dorsal fin longer than second (soft rayed); neither is fleshy. Sea raven, p. 330 First (spiny) dorsal fin much shorter than second (soft rayed); both thick and fleshy - Toadfish, p. 357 ' Although this flap suggests the adipose fin of a salmon in appearance, it is not actually an analogous structure, but simply a lobe of skin. 12 BULLETIN OF THE BUREAU OF FISHERIES Key A — Continued 18. Fishes wliich lie flat on one side, with both eyes on the other side, the upper side dark, the lower pale Flatfish tribe key, p. 472 Not lying flat on one side 19 19. Only one well-developed dorsal fin (this, however, may be preceded by isolated spines or rays) 20 Two or more separate and well-developed dorsal fins Key F, p. 13 20. Top of snout with several barbels or beards RockUngs(cod family in part) key, p. 385 No barbels or beard on top of snout 21 21. Jaws with very large canine tusks which project even when the mouth is closed Wolffish family key, p. 370 No large canine tusks 22 22. Dorsal fin soft-rayed, except that there may be a short spine at its forward margin Key G, p. 14 At least forward one-third of dorsal fin, if not whole length, spiny Key H, p. 15 Key B Fishes with tubular snouts (trom No. 3, p. 11). 1. Head horselike; trunk deep, narrowing abruptly to slender, prehensile tail; no caudal fin Seahorse, p. 177 Head roughly cylindrical; body very slender with no distinction into trunk and tail portions; caudal fin present 2 2. Snout no longer than dorsal fin; no ventral fin; caudal fin rounded Pipefish, p. 175 Snout more than six times as long as dorsal fin; ventral fins present; caudal fin forked Trumpetfish, p. 173 Key C Fishes with bills or swords (from No. 4, p. 11). 1. Both jaws elongated 4 Only one jaw elongated 2 2. Upper jaw elongated as a sword 3 Lower jaw elongated Halfbeak, p. 163 3. Sword sharp-edged; first dorsal fin shorter than the sword forward of eye; no ventral fins Swordfish, p. 221 Sword round-edged; dorsal fin nearly twice as long as sword Spearfish,' p. 227 4. Caudal fin well developed 5 No caudal fin; tip of tail is whiplike Snipe eel, p. 88 5. Several finlets behind dorsal and anal fins Needlefish, p. 164 No finlets behind dorsal and anal fins Silver gar, p. 161 Key D Fishes with well-developed fins, snouts of ordinary form, only one gill opening on each side, and the anal fin continuous with the caudal around the tip of the tail (from No. 13, p. 11.) 1. Only one dorsal fin ■- 2 Two separate dorsal fins, the first much higher than the second but shorter 6 2. Body band-shaped, the tail tapering to a whiplike tip Cutlasfish, p. 220 Body thick, eel-like; vertical fins continue around tip of taU in a broad band 3 3. Dorsal fin spiny from end to end 4 Dorsal fin soft rayed, at least for almost all its length 5 ' The sailflsh would also come under this heading should one ever be taken in the Gulf of Maine. The distinctions between it and the spearflsh are given under the account of the latter on page 228. FISHES OF THE GULF OF MAINE 13 Key D — Continued 4. Mouth large and strongly oblique; no ventral fins Wrymouth, p. 368 Mouth small and horizontal; small ventral fins Rook eel, p. 359 5. Without ventral fins Eel family key, p. 78 Small but distinct ventral fins, situated forward of the pectorals 6 6. Ventrals are situated behind the gill opening Eelpout family key, p. 376 Ventrals situated on the chin, well in front of the gill openings, and reduced to forked, barbel-like structures Cusk eel, p. 384 7. Ventral fins situated below point of origin of pectorals; eye very large — Grenadier family key, p. 467 Ventral fins situated behind tip of pectorals; eyes very small Chimeera, p. 73 Key E Bony fishes with two kinds of dorsal fins, i. e., one in front supported by rays, with a fleshy (adipose) fin behind it.' (From No. 14, p. 11.) 1. Rayed dorsal fin much longer than head, and spiny Lancetfish, p. 155 Rayed dorsal fin shorter than head, and soft-raj'ed 2 2. Jaws armed with long projecting saber-like fangs Viperfish, p. 153 Teeth small 3 3. Noticeable series of phosphorescent organs along each side 4 No phosphorescent organs 6 4. Mouth gapes back beyond eye Lanternfish family, p. 149 Mouth does not gape back as far as eye Pearlsides, p. 151 6. Tail deeply forked; nose pointed Smelt family key, p. 140 Tail nearly square or only slightly forked; nose rounded Salmon family key, p. 126 Key F Bony fishes with snouts of ordinary form, symmetrical tails, bodies not entirely encased in bony plate, caudal flna distinct from anal, and two or more well-developed dorsal fins, all of them supported by rays or spines (from No. 19, p. 12) . 1. All fins soft rayed; no spines Cod and silver hake families key, p. 385 First dorsal fin spiny; second soft-rayed 2 2. One or more small finlets between second dorsal and anal fins and the caudal 3 No such finlets 4 3. More than 3 each dorsal and anal finlets Mackerel family, p. 188 Two dorsal and two anal finlets Escolar, p. 220 Only one dorsal and one anal finlet Mackerel scad (pompano family in part), p. 232 4. Sides of head bony, with sharp spines or horns; head very broad 5 Sides of head have no spines or horns ; head not noticeably broad 6 5. Three lower rays of each pectoral fin separate from others, in the form of fleshy feelers; outline of tip of snout, as seen from above, concave; mouth small Sea robin family key, p. 345 Lower rays of pectorals not separate from others; outline of tip of snout convex, not concave; mouth very large Sculpin family key, p. 314 6. First spine of first dorsal fin very much stouter than others and can be locked erect by the second; no ventrals; skin very hard Triggerfish, p. 293 First dorsal spine not stouter than others; ventral fins well developed; skin soft 7 7. Ventrals more than twice as long as pectorals; caudal very small John Dory, p. 291 Ventrals no longer than pectorals; caudal fin large 8 8. Space between two dorsal fins is as long as the first dorsal; ventrals are situated be- • hind the middle of the pectorals 9 Little or no free space between the two dorsal fins; ventrals in front of middle of pectorals 10 ' The tileflsh (pp. 11 and 352) has a fleshy flap, simulating an adipose fln, on the back in front of the rayed dorsal fin. 14 BULLETIN OP THE BUREAU OF PISHEEIES Key F — Continued 9. Eyes large; mouth large and very oblique Silverside family key, p. 178 Eyes small," mouth very small and longitudinal Mullet, p. 182 10. Caudal peduncle extremely slender; caudal fin deeply forked Pompano family (in part) key, p. 229 Caudal peduncle moderately deep; caudal fin at most moderately forked 11 11. First (spiny) dorsal fin much lower than second (soft rayed) dorsal Bluefish, p. 237 First dorsal as high as second, or higher 12 12. Second dorsal fin not much longer than anal Sea-bass famDy (in part) key, p. 251 Second (soft rayed) dorsal about twice as long as anal fin Weakfish family key, p. 269 Key G Bony fishes with snouts of ordinary form, symmetrical tails, bodies not clad in bony plates, caudal fin distinct from the anal, neither canine tusks nor barbels on the top of the snout, and only one dorsal fin which is soft-rayed except that it may com- mence with one short spine (from No. 22, p. 12) . There is no adipose fin or flap either in front of the dorsal fin or behind itT 1. Tail deeply forked 2 Tail square or rounded 12 2. The whole of anal fin is behind the dorsal Herring tribe key, p. 90 Part or all of anal fin in front of rear margin of dorsal 3 3. Mouth gapes back beyond eye 4 Mouth does not gape beyond eye 6 4. Series of phosphorescent spots on each side 5 No phosphorescent spots or organs Anchovy, p. 124 5. Eye very large Lanternfish family, p. 149 Eye very small Cyclothone, p. 153 6. Eel-like in form Launce, p. 183 Not eel-like in form 7 7. Large ventral fins 8 Ventral fins wanting or very minute 10 8. Front portion of dorsal fin very high; body very deep Opah, p. 242 Dorsal fin not very high, tapers slightly from front to rear; general form slender, only about one-fifth as deep as long Pilotfish, p. 229 10. First dorsal rays very elongate with tiny ventral fins; deep and compressed in form Lookdown (adult), p. 236 First dorsal ray not elongate 11 11. Dorsal profile of head convex; forward portion of dorsal fin at least three times as high as rear part, narrowing abruptly; no ventral fins Butterfish family key, p. 245 Dorsal profile of head concave; dorsal fin tapers only slightly from front to rear; minute ventral fins Moonfish, p. 235 12. Dorsal fin preceded by one or more stout, isolated spines, with or without triangular fin membranes 13 No isolated spines in front of dorsal fin 15 13. Only one stout dorsal spine, situated over the eye; body very deep Filefish family key, p. 294 Several dorsal spines, all far behind the eye 14 14. Ventrals large, of ordinary form; caudal peduncle stout Barrelfish, p. 243 Each ventral consists of one very large stout spine, with or without a small fin membrane and one or two short weak rays; caudal peduncle very slender Stickleback family key, p. 166 FISHES OF THE GULF OF MAINE 15 Key G — Continued 15. No ventral fins, fishes capable of inflating themselves with air I Puffer and porcupine-fish families key, p. 297 Ventral fins present; can not inflate themselves with air 16 16. Dorsal fin runs whole length of back, from nape to base of caudal, which it joins; barbel on chin; form eel-like Cusk (cod family in part), p. 462 Dorsal fin occupies only one-third or less of back behind nape, leaving open space as long as fin between it and base of caudal; no barbel on chin; form not eel- like Mummichog family key, p. 155 Key H Fishes as in Key G, except that at least the forward one-third of the dorsal fin is spiny (from No. 22, p. 12). There is no adipose fin behind the rayed dorsal, nor fleshy flap in front of it. 1. Rear part of dorsal fin soft rayed 2 Whole length of dorsal fin spiny 7 2. Sides of head bony, with knobs or spines 3 No knobs or spines on sides of head 4 3. Sides of head with conical spines; spiny portion of dorsal fin at least as long as soft part; body laterally compressed Rockfish family, p. 304 Sides of head with low rounded knobs; spiny portion of dorsal fin considerably shorter than soft part; body tadpole-shaped-- Deep-sea sculpin (sculpin family in part), p. 329 4. Ventral fins much longer than pectorals; eye very large Big-eye, p. 261 Ventral fins no larger than pectorals; eye not very large 5 5. Pectorals pointed; body much compressed Sea bream family key, p. 263 Pectorals rounded; body not much compressed 6 6. Rear (soft) portion of dorsal fin nearly as long as anterior (spiny) part; anal much higher than long Sea bass (sea bass family in part), p. 251 Rear (soft) portion of dorsal fin less than half as long as spiny part; anal much longer than high Gunner family key, p. 280 7. Mouth strongly oblique; no ventral fins Wrymouth, p. 368 Mouth not strongly oblique; ventral fins present (very small in one species) Blenny family key, p. 359 THE LAMPREYS. CLASS MARSIPOBRANCHII Except for Amphioxus and its allies, the lampreys are the most primitive of vertebrates, their skeletons being cartilaginous and their skulls hardlj' differentiated from the vertebral column. They have no true jaws, no ribs, no shoulder or pelvic girdles, and no paired fins. They are eel-like in appearance, but are easily dis- tinguishable from the true eels and, indeed, from most of the true fishes by the pecuhar jawless sucking mouth situated at the tip of the snout, and from all Gulf of Maine eels by the absence of pectoral fins. THE HAGFISHES AND LAMPREYS. FAMILIES MYXINID.S; AND PETROMYZONID.a; These two groups are easily distinguished by the fact that the hags have but one gill opening on each side, one continuous fin on the back, and several barbels on the snout, whereas in the true lamprej's there are seven gill openings on each side, the fin on the back is separated into dorsal and caudal portions, and there are no barbels on the snout. 16 BULLETIN OF THE BUREAU OF FISHERIES 1. Hagflsh { Myxine glutinosa lAnnsdus) Jordan and Evermann, 1896-1900, p. 7. Description. — The hag, hke the lamprey, lacks paired fins and fin rays. Its skeleton is wholly cartilaginous, without bones, its mouth is similarly jawless, and its skin is scaleless. It is easily recognized by its eel-like form; by its single finfold (a fold of skin, not a true fin) running right around the tail and forward on the lower surface of the body with no division into dorsal, caudal, and anal fins; by the single gill pore on each side, just forward of the origin of the ventral finfold; by its lipless mouth, stellate in outline when closed; by the single nasal aperture at the tip of the snout; by its peculiar barbels or "tentacles," two flanking the mouth on either side and four surrounding the nostril ; and by the evertible tongue studded with rows of horny rasplike "teeth." We might also mention the series of mucus sacs on either side of the abdomen, and point out that the dorsal finfold originates two-tliirds and Fia. 2.— Hagflsh (Jfiiiineslufinosa). a, Aduit. !), Egg. AfterDean the ventral one-third the distance back from snout to tip of tail, ^^^th the vent piercing it. Hags vary in color, perhaps to correspond with the color of the bottom. They are grayish brown or reddish gray above, variously suffused, mottled, or piebald with darker or paler gray, brown, or bluish; below they are whitish or pale gray. They grow to a length of about 1 foot to a foot and a half. General range. — Arctic seas and both coasts of the north Atlantic; south in deep waters to the latitude of Cape Fear, N. C. (33° 50' N.). Occurrence in the Chilf of Maine. — The hag is only too common in the Gulf of Maine; perhaps it is not absent from any considerable area of smooth bottom. Thus it is abundant generally off the north end of Grand Manan, is reported from Passa- maquoddy Bay and from various localities near Eastport, is to be found offshore on muddy bottom all along the Maine coast, is caught at times in considerable numbers FISHES OF THE GULF OP MAINE 17 on the Boon Island — Isles of Shoals fisliing ground and about Jeffreys Ledge, where we found it plentiful enough in the spring of 1913 to have gutted 3 to 5 per cent of all the haddock in the gill nets. Fishermen report it as equally numerous in the deeper parts of Massachusetts Bay. On the offshore banks the hag is well known, and it has been trawled at various localities along the outer edge of the continental shelf off New England at depths of from 100 to 200 fathoms, and deeper. We ourselves took 11 large ones in one set of the Monaco deep-sea trap in 260 fathoms off Nantucket on July 9, 1908, and it has been taken in from 300 to 500 fathoms off Marthas Vineyard. Habits. — The hag is not a true parasite, as has sometimes been suggested, there being no reason to believe it ever attacks living, uninjured fish, but it is a scavenger. Judging from its habits during the brief time it survives in aquaria, it spends its time lying embedded in the clay or mud with the tip of the snout projecting, but it is an active swimmer. Probably it finds its food by its greatly specialized olfactory apparatus. So far as is known it feeds chiefly on fish, dead or disabled, though no doubt any other carrion would serve it equally well, were such available. It is best known for its troublesome habit of boring into the body cavities of hooked or gilled fishes, eating out first intestines and then the meat, finally to leave nothing but a bag of skin and bones, inside of which, or clinging to the sides of a fish it has just attacked, the hag itself is often hauled aboard. In fact, it is only in this way, or entangled on lines, that hags ordinarily are taken or seen. Being worthless itself, it is an unmitigated nuisance and a particularly loathsome one, owing to its habit of pouring out slime from its mucus sacs in quantity out of all proportion to its small size. One hag, it is said, can easily fill a 2-gallon bucket, nor do we think this is any exaggeration. The hag is at home only in comparatively low temperatures — cooler, probably, than 50° — and this confines it to depths of 15 to 20 fathoms or more in the Gulf of Maine in summer. Breeding habits. — The hag and its immediate relatives are hermaphrodites — the only regularly effective ones in the whole vertebrate series, except for a very few species of bony fishes. Its single unpaired sex organ first develops sperm in the rear, then eggs in the forward portion.' Further than this our knowledge of its breeding habits is still of the scantiest. Probably there is no definite spawning season, but eggs may be laid at any time of the year, for females near ripeness and others nearly spent have been recorded for various months, winter and spring as well as summer and autumn, and eggs have been taken in Norwegian waters from November to May. It has long been known that the eggs are large (up to 20 mm. in length), tough-shelled, and comparatively few (only 19 to 25 nearly ripe eggs having been counted in any one fish), and that they are very characteristic in appearance, for at each end they bear a cluster of barb or anchor tipped filaments (fig. 2b). Up until 1900 none had been found about which it could be asserted without hesitation that they had been laid naturally. In that year, however, Dean (1900) described hag eggs from the northwest part of Georges ' For an account of the sex organ of the hag see Schreiner (Biologisches Centralblatt, XXIV Band, Nr. 3, February, 1904, pp. 91-104). For a summary of earlier studies see Smitt (Scandinavian Fishes, 1892, p. 1205). 18 BULLETIN OP THE BUEEAU OF FISHEEIES Bank and from the south coast of Newfoundland. Jensen ° described others from the neighborhood of the Faroe Islands, and since then Huntsman has recorded them from the mouth of the Bay of Fundy and Hjort '" from Norway. The eggs are demersal and stick fast in clusters to some fixed object — in Jensen's case to a Bryo- zoan — both by their filaments and by slime threads. Newly hatched hags have never been seen, but inasmuch as the smallest yet described (about 2}^ inches long), probably not long out of the egg, already resembled the adult in external appearance there is no reason to suppose that the hag passes through a larval stage greatly different from the adult. The few egg finds thus far reported, being from 50 to 150 fathoms, point to rather deep water for the spa^vning of the hag. The Norwegian eggs mentioned by Hjort (taken in shrimp trawls) were on ooze bottom, but whether the hag invariably seeks this type of ground for breeding remains to be learned. I need only add that, to judge from Cunningham's experience with hags in aquaria, the females cease to feed with the approach of sexual maturity, as do so many other fishes. 2. Sea lamprey {Petromyzon marinus Linnaeus) Lamprey; Spotted lamprey; Lamper; Eel-sucker; Great sea lamprey Fig. 3.— Sea lamprey (Petromyzon marinus) Jordan and Evermann, 1896-1900, p. 10. Description. — Lampreys are very primitive vertebrates, eel-like in appearance, with soft, cartilaginous skeleton. They lack paired fins but have well developed dorsal and ventral finfolds. In the adult the jaws are so rudimentary that apparently they are wanting; the mouth is a longitudinal slit when closed, but when open forms an elliptical disk at the tip of the snout and is armed with many horny, hooked teeth arranged in numerous (11 to 12) rows, the innermost the largest. There are seven pairs of open gdl shts and two dorsal fin folds, whereas the hag has but one pore on each side and only one fin. The sea lamprey (the only member of its group known from our salt waters) can hardly be mistaken for any other fish, its eel-like appearance coupled with the jawless mouth sufficing to place it at a glance. Color. — In color the sea lamprey varies with locality, and perhaps wath age and season also. It is usually described as mottled above — hence the vernacular name "spotted lamprey" — and plain tinted below. While the ground color of the upper ' Videnskabelige Meddelelser tra Dansk naturhistorisk Forening i Kj^benhaven, 1800, p. 1. "Fishing experiments in Norwegian Fjords, by Johan Hjort and Knut Dahl. Report on Norwegian Fishery and Marine Investigations, Vol. I, 1900, No. 1, Chap. IV, p. 75. Kristiania. FISHES OF THE GULF OF MAINE 19 surface of the body of lampreys from Massachusetts Bay is perhaps most often ohve bro^\^l mottled with darker brown or black (the dark patches almost confluent), a plain bluish variety has been described, as have lampreys with the ground color yellowish brown, greenish, reddish, and bluish. Occasionally they are plain colored, but usually variously mottled. Perhaps the color of the bottom on which they live determines the color of lampreys as it does of so many other fishes. The lower sur- face is whitish, gray, or a pale shade of the same hue as the ground color of the back. During the breeding season lampreys (at least the landlocked form) are described as taking on more brilliant hues, the ground color between the dark spots turning bright yellow. SiBe. — The lamprey rarely grows to a length of 3 feet and a weight of 5 pounds or more. Usually, however, advdts, as they run up our rivers, are 2 to 2}^ feet long. General range. — Atlantic coasts of Europe and North America, from Labrador south to Florida in the western Atlantic. The lamprey spends most of its life in salt or brackish water, but ascends fresh-water rivers to spawn. Occurrence in the Gulf of Maine.— "No doubt the sea lamprey occurs along the whole coast line of the Gulf of Maine, for it is recorded in or at the mouths of numerous rivers and streams in Nova Scotia, New Brunswick, Maine, and Massa- chusetts, specifically in the St. John and Shubenacadie Elvers and from the St. Andrews region in salt water in the Bay of Fundy; from Eastport, Bucksport, Casco Bay, and the Presumpscott and Penobscot Rivers in Maine; from the Merrimac River; and from various stations in Massachusetts Bay, where it has been taken from time to time attached to driftwood and to the bottoms of boats as well as fastened to fishes. In olden times lampreys entered the Merrimac River in extraor- dinary numbers, but hard fishing has depleted their ranks sadly. Like other anadromous fishes, though they may seem plentiful enough when condensed in the narrow bounds of river banks, the stock probably is in no wise comparable with that of the commoner schooling fishes. Certainly they are not seen very often in the open sea. Probably at one time there was a run of lampreys in all the larger streams empt3dng into the Gulf of Maine, and they are still to be caught in the Merrimac, Kennebec, Penobscot, St. John, and Shubenacadie, and no doubt in sundry other rivers where we, personally, have no direct knowledge of them. Habits. — Large lampreys have long been known to run up New England rivers a Uttle earher in spring than do shad, possibly commencing to work upstream as early as the end of April. They appear regularly in the Merrimac in May, and are most abundant there in June, after which few if any enter. They go far upstream, even to the headwaters, where they spawn in June and July. A sea lamprey has been found to contain 236,000 ova. For the most complete survey of the hfe history of the lamprey we must turn to a landlocked race inhabiting certain lakes in the interior of New York and in Ontario. Briefly, it is as follows: " Such of the lampreys as approach '1 For an account of nest building and spawning, which are hardly germane to the present study since they do not take place in salt water, the reader is referred to Oage (The lake and brook lamprevs of New York, especially those of Cayuga and Seneca Lakes. The Wilder Quarter-Century Book, 1893, pp. 421-493, Pis. I-VII. Ithaca), HussakotI (Sea lampreys and their nests. American Museum Journal, 1913, Vol. 13, p. 323), and to Coventry (Breeding habits of the landlocked sea lamprey, Petramyzon marinus var. dorsatus Wilder. University of Toronto Studies, Biological Series, No. 20; Publications of the Ontario Fisheries Research Laboratory, 1922, No. 9, p. 133. Toronto). 20 BULLETIN OF THE BXJEEAU OF FISHEEIES sexual maturity run up from the lakes into small clear brooks to spawn in June. As they ripen, the two sexes become dissimilar in appearance, the males (and this is equally true of sea-run fish both in American and European rivers) developing a ridge along the back, the females a finlike crest between the vent and the caudal fin. They build nests of round stones, which they drag together with their suckerlike mouths, as has often been described and pictured in natural histories, and after spawning apparently most, if not all, die, for not only have they often been found dead but their intestines atrophy, they are attacked by fungus, and they become so debilitated that I'ecovery seems out of the question. In short, the old tradition that no lampreys return to the sea from the rivers they ascend seems well founded. The larvte are very different in appearance from the adults. They are blind and toothless, with mouths and fins of different shape. They continue in this state for a period estimated at 3 to 4 years, during most of which time they live in holes or burrow in the mud or sand, hiding under stones. Doctor Huntsman informs us, however, that they have been taken in tow nets in the Shubenacadie River in Nova Scotia. They subsist on minute organisms. At the end of this larval period, when they have grown to a length of 4 to 6 inches, they imdergo transformation to the adult form and structure, an event occupying about two months — August to October— and descend the streams of their nativity to the sea just before the water freezes in November or December, to live and grow there for one or two years or imtil they reach full size and sexual maturity. The larvse of the sea lamprey are very abimdant in the mud of flats near the mouths of small tributary streams of such river systems as the Delaware and Susquehanna, where lampreys breed abundantly, and they have been reported in the Shubenacadie (a stream emptying into the Bay of Fundy) and no doubt occur in the Merrimac and other Gvdf of Maine streams. Although lampreys spawn but once and then perish, their period of growth is so long that large ones, not yet mature, are to be found in salt water all the year round. Little is known of the habits of the lampreys while they five in the sea further than that the mode of life centers around a carnivorous nature. Judging from their landlocked relatives and from the occasions on wliich they have been found fastened to sea fish, they must be extremely destructive to the latter, wliich they attack b}^ " sucking on " \\dth their wonderfully effective mouths. Usually the lam- prey fastens to the side of its victim, where it rasps away imtil it tears through the skin or scales and is able to suck the blood. Its prey sucked dry, it abandons it for another. Probably lampreys are parasites and bloodsuckers, pure and simple, for we can not learn that anything but blood has been found in their stomachs, except fish eggs, of which lampreys are occasionally full.'^ Lamprej's have been found preying upon cod, haddock, and mackerel in Massachusetts Bay, even on basking sharks, and salmon, too, are said to be much aimoyed by them. When not clinging to anything they are strong, vigorous swimmers, progressing by an undulating motion in the horizontal plane, and they are said to be exceedingly aggressive in their attacks on other fishes. Occasionally they are found fast to driftwood, even to boats. " "The Fisheries and Fishery Industries of the United states," by George Brown Goode. Section I, 1884, p. 677. Wash- ington. FISHES OF THE GULF OF MAINE 21 How far offshore lampreys wander is not known. Probably, however, most of them remain in the coastal zone, if not in estuaries, and there is no evidence that they ever descend to anj^ considerable depth. A few were brought in from Georges and Browns Banks, however, during the early years of the Bureau of Fisheries.'^ Since lampreys never take the hook or are captured in nets except on rare occa- sions they are seldom seen in salt water; only when ruiming up our rivers are they familiar objects. In Europe, during the middle ages, lampreys were esteemed a great deUcacy — historians tell us Henry I of England died of a surfeit of them — and formerly, when they were much more plentiful than nowadays, considerable numbers were captured in the rivers of New England, particularly in the Connecticut and Merrimac Kivers. They were, indeed, regularly sought in the former until well into the last half of the past century, but for 40 years now the lamprey fishery has been hardly more than a memory except locally and in a small way for home consimiption. In the salt water of the Gulf of Maine the lamprey has never been of any commercial importance; the average fisherman might not see one in a Ufetime, nor is there any sale for the few picked up by chance. TRUE FISHES. CLASS PISCES Sharks and rays. Subclass Elasmobranchii The most obvious external character by which all sharks and rays are dis- tinguishable from the bony fishes is that there are five or more pairs of gill openings on either side of the neck, instead of onty one. In this they agree with the lamprej's, but it is a commonplace that their jaws and teeth are extremely well developed. Their skins are tough and leathery and studded -with denticles (placoid scales), which but remotely suggest orchnary scales and which are not homologous with the scales of bony fishes, for both dermis and epidermis take part in their formation, instead of the former alone. The teeth of the sharks and rays are essentially such placoid scales modified and simply embedded in the gums, not in the jaws. The fins are supported at their bases with segmented cartilaginous rods, and further out by numerous slender horny fibers, instead of by such rays or spines as are to be seen in the bony fishes. All the fins are covered with the same leathery skin that clothes the body. Among sharks the tail is uneven, with the vertebral column extending out into its upper lobe, but in most skates and rays it is whiphke, with no definite caudal fin. The torpedo (p. 6S) is an exception to this rule. The skeleton is for the most part cartilaginous, the skull far simpler than it is among the bony fishes, and the gills are attached throughout their lengths to the partitions between the gill openings instead of being free, while the rear portion of the digestive tract is modified into the so-called "spiral valve" by the develop- ment of a special fold from its lining layer. Sharks are usually looked upon as the most primitive of the true fishes. " Report of the Commissioner of Fish and Fisheries for 1879 (1882), pp. 811, 812, and 814. Washington. 22. BULLETIN OF THE BUKEAU OF FISHERIES SHARKS Sharks are always objects of interest, not only to fishermen and mariners but to seaside visitors generally, because of their evil appearance, their ferocity, the large size to which some of them grow, the destruction they wreak on fishermen's nets and lines as well as on the smaller fishes on which they prey, and the bad reputation certain kinds have earned, rightly or wrongly, as man-eaters. The Gulf of Maine is not particularly rich in sharks (compared with our south- em coasts, very poor indeed), for while the nimiber of species actually recorded there is considerable (indeed any high-seas shark might straggle thither) the little spiny dogfish alone is numerous in the sense in which this term is applied to the various commercial fishes. Only one of the larger species, the mackerel shark {Isurus punctatus), visits us in numbers sufficient for one to be fairly sure to see it during a summer's boating off the coast north of Cape Cod. With the larger sharks generally so scarce (the mackerel shark is weak-toothed and perfectly harmless to anything larger than the fishes on which it feeds), the danger of attacks on bathers is negligible. Indeed, not a single well-authenticated instance of the sort is on record" for the past SO years for the coast north of Cape Cod, though the beaches yearly are crowded with vacationists. As long as the white shark occasionally strays iato the Gidf, however (p. 40), it is always remotely possible that some summer we may be horrified by the news of such a tragedy as occurred on the New Jersey coast in July, 1916, when several persons were killed or injured, presumably by a shark of this species that was captured nearby a few days later.'^ Most Gulf of Maine sharks — certainly all the commoner ones — are viviparous, giving birth to young not only practically adult in structure but of relatively large size at birth. As sharks are of little commercial value in the Gulf of Maine (attempts to introduce the dogfish as a food fish having failed so far) they are an unmitigated nuisance to the fishermen because of their damage to nets and other gear. It is possible to identify all sharks so far known from the GuK — and tliis in- cludes all that are apt to occur there except as strays — by the size, structure, and relative locations of the fins, and by such tooth characters as may be seen at a glance at the open mouth or easily felt with the finger (after the shark is dead!). In the following descriptions of the several species we have attempted to present only such features as will tell what shark is at hand; for more minute par- ticulars we refer the reader to Garman's monograph (1913), which is not only the most authoritative work on this group of fishes, but in which almost all our species are beautifully pictured. 1* In 1830 — an event often quoted — one Joseph Blaney, fishing from a small boat in Massachusetts Bay off Swampscott, Mass., was attacked by some fish that was seen to overset and sink his boat and presumably devoiu-ed him, for neighboring fisher- men, who hastened to his rescue, found no trace of him. Whether his attacker was a large shark or, as we think more likely, a killer whale, is an open question. " Murphy and Nichols (The shark situation in the waters about New York. The Brooklyn Museum Quarterly, Vol. Ill, October, 1916, No. 4, pp. 145-160. Brooklyn) give a detailed account of this occurrence. FISHES OF THE GULF OF MAINE 23 KEY TO GULF OF MAINE SHARKS 1. Head hammer-shaped Hammerhead, p. 31 Head of ordinary shape; rounded or pointed nose 2 2. Only one dorsal fin; six gill slits on each side; body eel-shaped Eel shark, p. 24 Two dorsal fins (the second may be small but is always perfectly distinct) ; only 5 gill slits; body of ordinary shark form 3 3. Both dorsal fins have spines at their forward margins; no anal fin 4 Dorsal fins lack spines , 6 4. Rear margin of upper lobe of tail not notched ; a very common species Spiny dogfish, p. 44 Rear margin of upper lobe of tail notched near the tip 5 5. Dorsal spines so small they are hardly visible, though easily felt Portuguese shark (Centroscymnus c(Elolepis),p. 51 Both dorsal spines large Black dogfish, p. 53 6. There is no anal fin, the paired ventrals being the only fins on the ventral surface 7 Anal fin present 8 7. First dorsal fin situated about midway between pectorals and ventrals Greenland shark, p. 53 First dorsal far back as ventrals Bramble shark {Echinorhinus brucus), p. 55 8. No lateral keels on caudal peduncle (root of tail); upper lobe of caudal fin much longer than lower * 9 A longitudinal keel on either side of caudal peduncle; lower lobe of tail more nearly as large as upper, suggesting tail of a swordfish 14 9. Upper lobe of caudal fin nearly, if not quite, as long as head and body together Thresher, p. 32 Caudal fin less than half as long as head and body combined 10 10. Second dorsal at least half as high as first 11 Second dorsal less than half as high as first 12 11. Second dorsal considerably smaller than first; teeth small, blunt, and arranged like a pavement Smooth dogfish, p. 24 Second dorsal about as large as first; teeth narrow and pointed Sand shark, p. 34 12. Origin of first dorsal hardly behind pectorals; upper and lower teeth alike; skin spotted Tiger shark, p. 27 First dorsal originates well behind the pectoials; upper teeth broader than lower; skin not spotted 13 13. The first dorsal originates about over the inner corner of the pectorals when these are laid back; snout broadh' rounded Dusky shark," p. 29 First dorsal originates far behind inner corner of pectoral; snout long and pointed Blue shark, p. 28 14. Gill slits very long; first pair nearly meeting on throat; gills with rakers; teeth tiny.. Basking shark, p. 41 Gill sUts short, confined to sides of neck; no gill rakers; teeth large 15 15. Teeth broad, triangular, with serrate edges; second dorsal fin well forward of anal White shark, p. 39 Teeth slender, smooth-edged; second dorsal fin over or hardly in front of anal 16 16. First dorsal fin originates above axil (armpit) of pectoral Mackerel shark (Isurus punctatus), p. 36 First dorsal fin originates well behind the axil f pectoral Sharp-nosed mackerel shark (/. iigris), p. 38 i» The brown shark (Carcharinus milberti), very abundant west and south of Cape Cod but not yet known from the Oulf, is easily distinguished from its close relative, the dusky shark, by its very tall dorsal fin. 24 BULLETIN OF THE BUREAU OF FISHERIES THE EEL SHARKS. FAMILY CHLAMYDOSELACHID.E 3. Eel shark {CMamydoselachus anguineus Garman) Frilled shark; Snake shark; Sea serpent Jordan and Evermann, 1896-1900, p. 16. Garman, 1913, p. 14. Description. — The readiest field marks for this curious shark are the eel-Uke form of its body and tail, the fish being about fifteen times as long as deep; the fact that there is only one dorsal fin, situated far back over the anal but smaller than the latter; that there are six gill openings on a side instead of five; and that the mouth is more nearly terminal than in most sharks, with the snout hardly projecting Fio. 4.— Eel shark iChlamydoaelachtu anguineus). After Qoode and Bean beyond it. The pectorals, it may be added, are relatively small; the ventrals are larger and close in front of the anal. Size. — The few eel sharks so far recorded have been from 2 to 5 feet long. Color. — Uniform brown. General range. — Probably cosmopohtan in the deep waters of temperate and tropical oceans. This shark has been taken, on several occasions, in Sagami Ba}', Japan; also off New South Wales, Madeira, and Norway. Occurrence in the Gulf of Maine. — A curious eel-like fish found dead in a net near Pemaquid Point, Me., in 1880 " may have been an eel shark, and tliis is its only claim to mention here. It would not be surprising should it stray into our Gulf along the trough of the basin from the open Atlantic, for it is as likely to five off our coast as off any other, so widely separated are the localities of capture, listed above. " Described by Hanna (1883). FISHES OF THE GULF OF MAINE 25 THE SMOOTH DOGFISHES. FAMILY GALEORHINID^ These are rather small sharks (17 known species) with two dorsal fins, the first large and the second usually much smaller, without spines. The upper lobe of the tail is much longer than the lower, anal fins are present, and the teeth are flat and pavementhke. Except for the teeth they closely resemble the requiem sharks (family Carcharinidse, p. 27). 4. Smooth dogfish {Galeorhinus Isevis Vaimont) Grayfish; Smooth dog; Smooth hound; Switch-tail; Whtpper-tail Jordan and Evermann (Mustelus canis Mitchill), 1896-1900, p. 29. Garman, 1913, p. 176. Description. — The smooth dog is easily identified by the presence of two large spineless dorsal fins, the first larger than the second, combined with an anal as well as the paired ventral fins on the lower surface; a tail of typical shark outline — that is, the upper lobe longer than the lower but not excessively elongated — and with flat granular teeth. So different, indeed, are the teeth from the cutting teeth Fio. 5.— Smooth dogfish (Galeorhinus Isem) of all our other sharks, that a glance at the mouth is enough to separate this species from the young of any larger Gulf of Maine shark. In form this httle shark is slender, flattened below, with tapering but blunt snout. Its first dorsal originates nearly over the hind angle of the pectorals and is decidedly larger than the second. The latter, in turn, is about twice as large as the anal, over which it stands. The hind margin of the upper lobe of the caudal is deeply notched near the tip ; the lower caudal lobe is very small. Size. — ^Adult smooth dogs average about 2 to 3 feet in length, but they have been taken up to 5 feet in length. Color. — Light gray above ; paler gray below. General range. — Cape Cod to Cuba in American waters; also off the coasts of southern Europe. Occurrence in the Gulf of Maine. — The smooth dog is extremely abundant west and south of Cape Cod. In Long Island waters, for example, it is one of the commonest and most generally distributed of fishes from June until November, and it aboimds equally throughout the summer and early autumn in the Woods Hole region on all kinds of bottom. This, however, is the most easterly outpost for its presence in any numbers, for though it has been reported from Provincetown, 26 BULLETIN OF THE BUREAU OF FISHERIES from various localities within Massachusetts Bay, and even from as far north as St. Andrews in the Bay of Fundy, where one was caught in July, 1913, it occurs only as a southern straggler in the GuK of Maine, and that so rarely that neither of the authors has ever seen it north of Cape Cod. So far as known its occasional incur- sions into the Gulf are sporadic — at least they have not been correlated with unusu- ally warm summers or ■with the presence of other southern fishes. On the outer part of the continental shelf Nantucket Shoals must be regarded as the easterly limit of its regular occurrence, for it is not recorded nor reported by fishermen from either Georges or Browns Banks, nor was it detected there by the representatives of the Bureau of Fisheries during the trawling investigations of the years 1912 and 1913 (p. 9). The smooth dog is most familiar as a shore fish and a bottom swimmer, com- monly entering shoal harbors and bays, nor is it known to descend to any consider- able depth. Food. — The food of the smooth dog consists chiefly of the larger Crustacea, and it is perhaps the most relentless enemy of the lobster, which had been eaten by no less than 16 per cent of the fish examined by Field (1907). Large crabs are like- wise an important article in its diet, as are the smaller fishes. Field estimated that in Buzzards Bay 100,000 smooth dogfish would annually devour over 600,000 lobsters, 90,000 to 100,000 fish of one kind or another (menhaden and tautog are the species most often found in dogfish stomachs) , and a couple of million crabs. While these figures are to be taken only as broadly suggestive, they are based on a suf- ficient number of observations of the stomach contents to serve as a general indi- cation of the destructiveness of dogfish. They also feed on squid, especially in spring, and while they do not regularly take moUusks, razor clams have been found in the stomachs of several at Woods Hole. When kept in captivity they are con- stantly on the move, searching the bottom for food, which they find chiefly by the sense of smell though their sight is also keen.'' Any crab that may be offered is soon found, seized, shaken to and fro, and eaten, and with packs of these sea hounds hunting over every square foot of our southern bays and sounds it is a wonder any of the larger Crustacea escape when dogfish are abxmdant. Field also made the interesting observation that the smooth dogs never molested healthy and active menhaden but soon devoured any sick or injured fish that might be in the same tank with them. Breeding habits. — Not being a characteristic Gulf of Maine fish we need merely note of its breeding habits that it is viviparous,'* giving birth to from 4 to 12 young at a litter, the pups being about a foot long and practically of adult form when born ; and that in the Woods Hole region females containing eggs and embryos at various stages in development are to be found throughout the summer. How many litters of young are produced by any one female during a year is still to be learned. " The senses of this shark have been studied by Parker (Bulletin, V. S. Bureau of Fisheries, Vol. XXIX, 1909 (1911), pp. 43-57) and by Sheldon (Journal of Comparative Neurology and Psychology, vol. 19, 1909, No. 3, p. 273). "In the report of the Massachusetts Commissioners of Fish and Game for 1905 it is erroneously said to be oviparous, apparently being confused with the European dogfish, ScyWum cankula. FISHES OF THE GULF OF MAINE 27 REQUIEM SHARKS. FAMILY CARCHARINID^ This family, containing a large number of species in tropical and temperate seas, is characterized by a head of normal shape, tail with the upper lobe much larger than the lower but not greatly elongate, two spineless dorsal fins, the first usually much larger than the second and situated over the space between the pectorals and the ventrals, a caudal peduncle lacking lateral keels, and sharp teeth. 5. Tiger shark {Galeocerdo arcticus Faber) Jordan and Evermann (G. tigrinus MuUer and Henle), 1896-1900, p. 32. Garman, 1913, p. 148. Description. — The tiger shark is characterized among the "smooth" (spineless) sharks by the fact that it has an anal as well as ventral fins, that the upper lobe of the tail is much larger than the lower, that the second dorsal fin is very much smaller than the first, and that the latter originates little, if any, behind the "armpit" of the pectoral. The only Gulf of Maine shark with which it might be confused is the dusky shark (p. 29), but it is easily separable from the latter by the more forward position of the first doreal fin and by the fact that it is spotted instead of plain colored. I may also note that its teeth are large and alike in both jaws. The body is slender, rather heavy forward of the pectorals, and tapering toward the tail. The head is large, very short, and broad. The snout is rounded FiQ. 6.— Tiger shark (attleocerdo arcUcus) (not pointed) and the mouth is very broad, occupying nearly two-thirds of the width of the snout. The first dorsal is high, triangular, and nearly as large as the pectoral, while the second dorsal is hardly one-third to one-fourth as high as the first and stands over the anal, which is of about equal size. The lower tail lobe is almost half as long as the upper, the rear margin of which is notched near the tip. Color. — Young tiger sharks are light brown, more or less spotted and barred with darker brown. These markings fade with advancing age until adults are nearly plain colored. Size. — This is one of the largest sharks, frequently being 12 to 15 and occa- sionally as much as 30 feet in length, though such a size is altogether exceptional. Most specimens caught north of the Carolinas are small. General range. — Cosmopolitan in the warmer waters of all oceans, whence it strays northward as far as Cape Cod on the American coast of the Atlantic. Occurrence in the Gulf of Maine. — Every year a few young tiger sharks are taken in the fish traps in the Woods Hole region, where, according to the records of the Bm-eau of Fisheries, it is the latest shark to arrive, rarely being seen before August 28 BULLETIN or THE BUREAU OF FISHERIES or after October. The specimens captured there usually have been about 5 feet long, and very rarely does a full-grown tiger shark stray so far from its tropical home. But, curiously enough, one at least of the several specimens recorded from Provincetown, its furthest known outpost and the only locality where it has been captured in the GuK of Maine, must have been of good size, for its stomach contained a whole full-groAvn swordfish. Habits. — This slender, active, and voracious shark, with wide jaws and powerful teeth, is an inhabitant of the high seas, preying upon the large sea turtles, other sharks, fish, and occasionally on invertebrates such as horseshoe crabs, crabs, conchs, whelks, etc. Remnants of squeteague, mackerel, hake, scup, menhaden, goosefish, and dogfish all have been found in stomachs of tiger sharks taken at Woods Hole.=° In the West Indies it is much dreaded, whether or not with good cause. So seldom does this species round Cape Cod (in fact none has been reported east or north of the cape for many years) that the chance of running across one in the Gulf of Maine is extremely remote. It has never been recorded from the offshore banks. 6. Great blue shark {Galeus glaucus Linnaeus) Blue Shark Jordan and Evermann (Prionace glauca Linnseus), 1896-1900, p. 33. Garman, 1913, p. 145. Description. — The blue shark is slender bodied, thickest at about its mid- length, and tapering thence toward the head and tail (that is, the shape usually named "fusiform"), its long pointed snout separating it at a glance from the blunt- nosed tiger. The first dorsal is of moderate size, standing well behind the middle of the space between pectorals and ventrals. The pectorals are very long, their tips reaching as far back as the first dorsal, and their very narrow and pointed outlines, combined with the location of the first dorsal and the pointed snout, give it an aspect very different from that of the dusky shark, which resembles it in the relative sizes of the fins. The second dorsal is less than half as high as the first — about equal to the anal over which it stands. The lower lobe of the tail is only one-third as long as the upper. The latter is notched near the tip, and both tail lobes are sharp pointed. The teeth of the blue shark are very characteristic, being large and serrate, each series forming a continuous cutting edge. Those of the upper jaw are broadly triangular with curved tips, while the lower teeth are narrower, pointed, and stand more erect. Size. — The blue shark grows to a length of about 12 feet. Color. — The color varies from grayish to light or bright steel blue, or even to bluish black above. Below it is dirty white. General range. — Cosmopolitan in the warmer parts of all oceans. On the northeastern coast of North America it is taken from time to time at Woods Hole, '" Bell and Nichols {Copeia, No. 92, Mar. 15, 1921, pp. 17-20) list the stomach contents of a large number of tiger sharks caught off Morehead City, N, C. FISHES OF THE GUU OF MAINE 29 where it is one of the rarer sharks, and at Nantucket. While only a stray in the Gulf of Maine, it must visit the outer coasts of Nova Scotia in some numbers every summer, for Harry Piers, of the Provincial Museum, Hahfax, informs us that there are three specimens in the museum — one of them 10 feet 5 inches long — taken near Hahfax. He also reports a fourth taken there in 1895, and writes that this shark was "plentiful at entrance to Halifax Harbor about 25 August, 1920; first seen about 15 August; last seen 23 September." Cornish ^^ also saw two specimens at Canso, Nova Scotia, but whether the "blue dogs" described to him by local fisher- men as common on the neighboring fishing banks actually are this shark seems doubtful. On the European side of the Atlantic the blue shark is not uncommon in summer around the south coasts of Great Britain, and has been taken casually as far north as southern Norway. Occurrence in the Gulf of Maine. — The claim of this species to mention here rests on a single specimen from Massachusetts Bay recorded by Garman (1913), but being comparatively so common off Nova Scotia it is to be expected in the Gulf Fig. 7. — Great blue shark {Gaieus glaucus) any summer. It may be noted in passing that it is viviparous, and that Nichols and Murphy " have given a graphic account of it as it is met with by whalers on the high seas. 7. Dusky shark. {CarcharMnus oiscurus heSueur) ■fe-^ Shovelnose Jordan and Evermann, 1896-1900, p. 3.5. Garman, 1913, p. 130. Description. — In the dusky shark (a moderately stout-bodied species) the second dorsal is not over one-half as high as the fii-st. . The latter stands well back of the pectorals, but, being nearer these than to the ventrals, is relatively further forward than in the blue shark and further back than in the tiger shark. The rear margin of the first dorsal is deeply concave; the pectorals are relatively long and narrow (twice as long as broad) and reach back as far as the rear edge of the first dorsal. The second dorsal is even smaller than the anal, over which it stands. The tail is long, occupying more than one-fourth of the total length of the shark, >i Further Contributions to Canadian Biology, 1902-1905 (1907), p. 81. In 39th Annual Report of the Department of Marine and Fisheries, 1906, Fisheries Branch. Ottawa. " Brooklyn Museum Science Bulletin, vol. 3, No. 1, 1916, p. 9. Brooklyn. 30 BULLETIN OF THE BUREAU OF FISHEHIES but its lower lobe is relatively shorter than in either tiger or blue shark. The dusky shark is further distinguished from the latter by its blunt rounded nose and broad flat head. The upper teeth are broad, triangular, serrate, and with concave outer edges; the lower teeth are narrower, more pointed, with broad bases, and stand more erect. Size. — This shark occasionally reaches a length of 14 feet, but the larger specimens caught in the traps are usually only 6 to 9 feet long. The relation of length to weight may be judged from the fact that one 11 feetij inches in length weighed 650 pounds. Color. — Gray brown above; whitish below. It is said that this shark is some- times blue above. ' General range. — Middle Atlantic; from North Carolina to Portland, Me., on the coast of North America. Occurrence in the Gulf of Maine. — Like several other sharks the shovelnose is sufficiently plentiful all along the shores of southern New England, as far east as Fig. 8. — Dusky shark ( CarchaThinu^ obscurus) Cape Cod, throughout summer and early autunm, to be well known to the local fishermen. At Woods Hole, for example, it is very common, but it rarely strays into the colder waters beyond the cape. The localities within the Gulf of Maine where it has been definitely recorded are Crab Ledge off Chatham, Nahant, Massa- chusetts Bay, and Cod Ledge near Cape Elizabeth (the most northerly occurrence yet known), where one was caught in 1864 by Capt. B. J. Willard. So rare are these stragglers that neither of the writers has ever seen one in the Gulf. In short, it has no place in the fauna of the latter except as a stray. Neither recorded cap- ture nor fishermen's report credits it to Georges or to Browns Bank. Food. — The shovelnose is a bottom swimmer, feeding chiefly on fish and squid but also eating the larger Crustacea. Gunners, menhaden, scup, skates, and silver hake have been found in specimens caught at Woods Hole. It is harmless to human l)eings. THE HAMMER-HEADED SHARKS. FAMILY CESTRACIONTID.S; The peculiar shape of the head, described below, sufficiently characterizes the only Gulf of Maine representative of tliis familj^, which otherwise resembles the requiem sharks (p. 27). FISHES OF THE GULF OF MAINE 31 8. Hammerhead shark {Cestracion zygsena Linn.-Bus) Jordan and Evermann {Sphyrna zygsnia Linnseus), 1896-1900, p. 45. Garman, 1913, p. 157. Description. — The bizarre outline of the head of the hammerhead, easier drawn than described, has been so widely heralded that probably everyone at all concerned with fishes is perfectly familiar with it. It can not possibly be confused with that of any other fish. The eyes stand at either edge of the "hammer"; the first dorsal fin originates slightly behind the "armpit" of the pectoral, is con- siderably larger than the latter, and is much higher than long; the very small second dorsal is hardly one-fifth as high as the first; the upper lobe of the tail is notably long (about one-third as long as the body of the fish) and deeply notched near the tip, the lower lobe hardly one-half as long as the upper. Size. — The hammerhead is one of the larger sharks, growing to a length of 15 feet or more. Color. — Gray to ashy brown above; paler brown to dirty white below. General range. — A warm-water species, cosmopolitan in tropical seas north- ward to the Gulf of Maine in the western North Atlantic, and to British waters in the eastern North Atlantic. FiQ. 9. — Hammerhead shark ( Cestracion zygxna) Occurrence in the Gulf of Maine. — The hammerhead, like most of its tropical relatives, finds Cape Cod and the cool water that it meets when it strays beyond that natural boundary the eastern and northern limit to its regular annual occur- rence. In the Woods Hole region, only a few miles west of the cape, it is caught from time to time in the fish traps from July to October almost every year. So far, however, the only definite reports of it in the Gulf of Maine with which we are acquainted are from Chatham and Provincetown, the latter its most northerly record on the American coast; nor is it likely that the hammerhead is more common in the Gulf than these few records suggest, for so easily recognized is it among sharks that it is far more apt to bo reported than are the various tropical species of more conventional appearance. It would not be surprising to see it on Georges or Browns Bank, though no rumor of its presence there has reached us. With the hammerhead, as with many other tropical fishes, the examples that visit the shores of New England are usually small. At Woods Hole about 4 feet is the commonest length and 6 to 8 feet the maximum. In 1805, however, a speci- men 11 feet long was netted at Riverhead, Long Island, N. Y., and the fact that 102274—251 3 32 BULLETIN OF THE BUREAU OF FISHEBIES this specimen contained parts of a man in its stomach is chiefly responsible for the bad reputation of the hammerhead. Habits. — The hammerhead is pehigic in habit, often swimming with dorsal and caudal fins above the surface. It feeds chiefly on fish and squids but is also known to eat crabs and even barnacles. It is viviparous. Thirty-seven embryos have been taken from the oviducts of a female 11 feet long, and probably such specimens as wander north of the Chesapeake Capes give birth to their young in summer, for specimens as small as 1}4 feet long have been taken at Woods Hole in Jul}- and August. THE THRESHER SHARKS. FAMILY VULPECULID.5; The only representative of this famih- (the well-known thresher) is peculiar among sharks for its enormously elongate tail. Its closest affinities otherwise are with the mackerel sharks (p. 35). 9. Thresher {Vtdpecula marina Vahnont) Thraser; Swiveltail; Swingletail; Fox shark Jordan and Evermann (Alopias vulpes Gmelin), 1896-1900, p. 45. Garman, 1913, p. 30. Fig. 10. — Thresher shark ( Vulpccula marina) Description. — The thresher is as easily distinguished by its long tail as the hammerhead is by its head, the upper caudal lobe being about as long as the head and body of the fish together, curved much like the blade of an ordinary scythe, and notched near the tip, whereas the lower lobe is hardly longer than the anal fin. It need merely be pointed out in addition that the first dorsal (of moderate size and about as high as long) stands about midway between pectoral and ventral, that the second dorsal and the anal are very small, the pectoral is very long and sickle shaped, and that the thresher is a stout-bodied shark with short snout, blunt, rounded nose, and small triangular teeth. Size. — The thresher grows to a length of about 20 feet or more, fish as largo as 16 feet in length having several times been taken at Woods Hole. One of 13 feet has been found to weigh about 400 pounds. Color. — Dark lead brown to nearly black above; white below, except that the lower sides of the pectorals are leaden in hue. General range. — An inhabitant of all warm seas, especially numerous in the Mediterranean and temperate Atlantic. FISHES OF THE GULF OF MAINE 33 Occurrence in the Gulf of Maine. — The most northerly locahty on the east coast of the United States where the thresher can be called fairly abundant is off Block Island, where, say Nichols and Murphy,^^ it is the commonest large shark, appear- ing in May, most plentiful in June, and remaining until late in the fall. At Woods Hole, too, it has occasionally been taken in the fish traps from April until late in the autumn. Specimens as large as 20 feet in length have been caught there — three fish of 16 feet each in one trap in a single morning. Although only two specimens have been reported at Nantucket, the thresher evidently enters the Gulf of Maine more often than do most of its tropical relatives (e. g., the blue shark) for it has been recorded repeatedly on the coasts of Maine and Massa- chusetts — at Provincetown, Massachusetts Bay, Boston Hai'bor, Nahant, off Monhegan, east of Matinicus, off Penobscot Bay where a specimen estimated to weigh 500 pounds was caught in 1911, and off Eastport. It is said to have been taken — even to have been common — in the past in the Bay of Fundy, though there is no recent record of it there, and it has been reported entangled in nets off the Nova Scotian coast and even from the Gulf of St. Lawrence. To these records we can add that of several large threshers seen leaping near the Grampus as she saUed through Pollock Rip on August 4, 1913. In fact, next to the mackerel sharks (p. 35) the thresher is no doubt the commonest large pelagic shark in the Gulf. No doubt it also occurs in the mackerel season on Georges and Browns Banks, though we find no definite record of it there. The thresher is to be expected in our waters only in the spring, summer, and autumn; in the cold season it alto- gether deserts the northern coasts for warmer seas. Food and habits. — The tale that the thresher leagues with the swordfish to attack whales is time honored, but it seems that it must be relegated to the category of myth, for few, if any, experienced whalemen can be found to credit it (except in yarns spun to entertain and awe landlubbers!), and so weak toothed is this shark that the second part of the story — that it makes a meal on its huge victim — is an impossibility. In actual fact the thresher feeds chiefly, if not exclusively, on such schooling fishes as mackerel, menhaden, herring (of which it destroys great numbers) , and, in European waters, pilchard. A pair of threshers often work in concert "herd- ing" a school of fish, and it is to frighten its prey together that its enormously long, flail-like tail is Employed. Allen ^* gives an interesting eyewitness account of a thresher pursuing and striking a single small fish with its tail. It is, we may add, perfectly harmless to human beings. Commercial importance. — In the Gulf of Maine the thresher is not common enough to be of any importance to fishermen one way or another, or to play a practical role of any moment among the smaller fish. Further south, however, and wherever it is numerous in the Atlantic, it makes itself a great pest, tangling and tearing mackerel nets as well as destroying and chasing away the more valuable fishes on which it feeds. " Brooklyn Museum Science Bulletin, vol. 3, No. 1, 1916, pp. 1-34, pis. 1-3. Brooklyn. '< Science, New Series, Vol. LVIII, No. 1489, July, 1923, pp. 31-32. 34 BULLETIN OF THE BUREAU OF FISHERIES THE SAND SHARKS. FAMILY CARCHARHDiE In the sand sharks the two dorsal fins are spineless and nearly equal in size, the upper lobe of the tail is much larger than the lower, there are no keels on the caudal peduncle, and the teeth are very slender and pointed. 10. Sand shark {Carcharias taurus Rafinesque) Shovelnose; Dogfish shark; Blue dog; Little mackerel shark; Ground shark Jordan and Evermann {Carcharias littoralis Mitchill), 1896-1900, p. 46. Garman, 1913, p. 25. Description. — The large size of the second dorsal and anal fins (which are about equal to the first dorsal instead of much smaller) is of itself enough to distinguish this species from all other GuK of Maine sharks. The first dorsal fin being located but little in front of the ventrals, the trunk seems crowded with fins of equal size — a useful field mark for this species. We may also point out that the pectoral fins are not much larger than the other fins — triangular rather than sickle shaped ; that the upper lobe of the tail is nearly one-third as long as head and body together and notched near the tip, with the lower lobe about one-fifth as long as the upper; and Fig. 11.— Sand shark (Carcharias taurus) that the head is flat, the nose short and blunt at the tip. The teeth of the sand shark (they are alike in both jaws) are likewise diagnostic, being long, narrow, and pointed, with a spur at either side near the base, and smooth-edged. Size. — Adult sand sharks are usually about 4 to 5 feet long, often a foot or more longer, and rarely taken up to 8 or 9 feet." They have been reported up to 12 feet long, but this is so much longer than the general i-un as to refise the question whether these monsters were actually sand sharks and not some other species. Color. — The ground color is gray, darker above, lighter below, indistinctly spotted with darker brown, and the edges of the fins are sometimes edged with black. General range. — Coastal waters of the United States from Maine to North Carolina. Occurrence in the Gulf of Maine. — The sand shark is the commonest of all its tribe (except the smooth and spiny dogfishes) at the westerly entrance to the Gulf of Maine. It is very plentiful at Woods Hole from June to November and is to be found everywhere in that region in shoal waters, even coming up to the wharves. At Nantucket, too, it is so abundant that shark fishing, with the sand shark as the chief objective, is a popular sport, and although we find it far less abundant once we " Sherwood (Copeia, Nov. 15, 1921, No. 100, p. 7") records one of 8 feet 10 inches, caught at Clinton, Conn. FISHES OF THE GXJLF OF MAINE 35 pass the southern elbow of Cape Cod, it is more often seen and taken in the Gulf of Maine than is any other large shark except the mackerel shark (p. 36) or perhaps the thresher. There is well-established record of its presence at Monomoy, North Truro, Provincetown (where it has been caught often enough to have received the local name of dogfish shark, appropriate because of its small size), Cohasset in Massachusetts Bay (where the senior author caught one about 4 feet long years ago) , in Boston Bay, at Lynn, at the mouth of Casco Bay, and even near St. Andrews in the Bay of Fundy — its most northerly outpost — where a stray specimen was taken in a weir in 1913. Probably, were all the sand sharks that entangle themselves in nets reported, we would find that it ranges northward as far as Casco Bay every summer and in much greater numbers than the actual published record would suggest. Any "shovelnose" reported from northern New England would probably belong to this species; and no doubt it is represented among the "ground sharks" taken by fishermen on Georges Bank, though definite information is lacking on this point. Habits and food. — This shark, in the warm months at least, swims chiefly near the bottom in shoal water, often coming right up on the beaches almost to tide mark and even entering the mouths of rivers. Over certain bars, however, it often comes to the surface, where it may be seen moving slowly to and fro with its dorsal and tail fins projecting above the surface. It captures great numbers of small fish, which are its chief diet, particularly menhaden, cunners, mackerel, skates, silver hake, flounders, alewives, butterflsh, and — south of Cape Cod — scup, weakfish, and bonito. It also eats lobsters, crabs, and squid. Although comparatively sluggish in habit, as sharks go, sand sharks have been seen surrounding and devouring schools of bluefish, and have even been known to attack nets full of blueflsh, which gives a measure of their voracity. There is no record or even well-grounded rumor that this shark ever attacks human beings. Indeed, it is looked upon merely as a harmless nuisance wherever it is common enough to be familiar. So far as the Gulf of Maine and, indeed, the southern coast of New England as a whole are concerned, the sand shark occurs only as a summer visitor, moving away either southward or into deep waters during the cold season. Breeding habits. — Nothing is definitely known of its breeding habits. Females with unripe eggs have been taken at Woods Hole in July. Commercial value. — This shark has no commercial value except the negative one of damaging nets, but so readily does it bite a hook that it is of some importance as an object of sport, though hardly so in the Gulf of Maine, where it is never plentiful enough to be Avorth fishing for. THE MACKEREL SHARKS. FAMILY ISURID.« This group of sharks is easily recognizable by the fact that the tail is very firm and lunate in outline with the lower lobe but little smaller than the upper, suggest- ing a swordfish's tail, and that there is a prominent keel on either side of the caudal peduncle. The dorsal fins are spineless. 36 BULLETIN OF THE BUREAU OF FISHERIES 11. Mackerel shark {Isuriis punctatus StoTer) Blue shark; Porbeagle Jordan and Evermann {Lamna cornuhica Gmelin), 1896-1900, p. 49. Garman, 1913, p. 36. Descri-ption. — The mackerel sharks (this and the two species following) are easily told from all the sharks so far mentioned by the shape of the tail, for while its lower lobe is sharklike, somewhat smaller than the upper, the difference is slight, the tail being almost evenly forked, with the upper lobe directed so sharply upward, the lower downward, that the tail as a whole is crescentic and much broader than long. In fact it recalls the tails of such pelagic bony fishes as the mackerel tribe or the swordfish in outline, likewise in its firm texture. More precise if less obvious a character is that the root of the tail bears a well marked longitudinal ridge or keel on either side, a feature shared by the white and basking sharks (pp. 39 and 41) . This is a stout, heavy-shouldered shark, tapering in front to a sharply pointed snout and behind to a very slim tail root. Its dorsal and pectoral fins are very large; the former, originating over the armpit of the pectoral, is triangular and about as high as long; the latter, broad-based but tapering sicldelike to a narrow tip. Fig. 12. — Mackerel shark {Isutu^ punctatus). After Garman is only about half as broad as long. The second dorsal and anal fins are very small indeed, and the ventrals but little larger. The second doi-sal stands over the anal. The positions of the dorsal fins are the readiest field mark to distinguish this species from the sharp-nosed mackerel shark (p. 38). The teeth are alike in the two jaws- small, slender, pointed, smooth-edged, and without spurs on the sides — and their structure differentiates this shark from the European porbeagle {Isurus nasus), which it otherwise resembles closely but in which the teeth bear a sharp denticle on either side at the base of the cusp. Size. — The larger mackerel sharks are usually about 8 to 10 feet long, growing to an extreme length of about 12 feet. Color. — The upper parts are dark bluish gi'ay to bluish brown, changing abruptly to white below. According to Garman the dorsal, pectoral, and tail fins are tipped with black, there is a black area in the armpit of the pectoral followed by a white space on the fin and body, and there is a large and very noticeable black spot on the outer half of the pectoral, which is one of the distinguishing features of this species. FISHES OF THE GULF OF MAINE 37 General range. — North Atlantic and Pacific Oceans. Closely allied to the ommon porbeagle (Isurus nasus Bonaterre) of British seas. Occurrence in the Gulf of Maine. — From the days of the earliest settlement it has been known that stout-shouldered, surface-swimming sharks of moderate size and with "mackerel" tails are tolerably common in the Gulf of Maine, universally referred to by the fishing population as "mackerel sharks." During the first half of the last century only one such shark species was recognized in our waters, but more recent researches have proved that there are actually two — the present one and the next — readily separable by the position of the first dorsal fin relative to the pectorals and of the second dorsal relative to the anal, but so much alike in general appearance that it is usually impossible to determine without actually exam- ining the specimens to which species many of the records actually belong. However, since /. punctatus is the more northerly of the pair, and since far more specimens of it than of /. tigris have actually come to hand, probably most of the mackerel sharks that fisherman so often see swimming lazily on the surface off the shores of Northern New England belong here. Although these sharks are far more often seen than captured, we have definite record of the common mackerel shark at Provincetown, in Massachusetts Bay, off Cape Ann, and at various localities along the Maine coast — e. g., off Cape Eliza- beth, in Casco Bay, off Monhegan, and even Passamaquoddy Bay in the Bay of Fundy, where, however. Huntsman (1922a) records but a single specimen. During our Grampus cruises we have seen many mackerel sharks, particularly between Cape Ann and the Isles of Shoals, and off Monhegan Island. This shark likewise ranges northward along the Nova Scotian coast and into the Gulf of St. Lawrence. It may, in fact, be described as common, if not abundant, and to be expected anywhere in the Gulf of Maine during the summer. In winter it apparently departs (no doubt for warmer seas), and it is during its southward journey throughout autumn and up to the end of November that mackerel sharks are commonest in the Woods Hole region, while at Nantucket they (or the next species) are commonest in spring when they are taken in the mackerel drift nets. As yet our knowledge of the migrations of this shark into and out of the Gulf of Maine is of the haziest. Certainly, however, it visits us in greater or less number annually, and is most numerous when mackerel are plentiful. Habits. — The whole mackerel-shark tribe, as contrasted with the ground sharks, are strong, active swimmers, leading a pelagic life near the surface of the high seas, wandering about over the ocean in pursuit of the fishes on which they prey, and often uniting in small companies, though they can hardly be called gregarious. Like swordfish they spend much time at the surface on calm days, when their triangular back fins, followed by the tip of the caudal fin (the bluntness of the former and the wavy track of the latter identify the shark as such) may often be seen cutting through the water. Again and again we have sailed up on sharks probably of this species, only to see them sound, just out of harpoon range, plainly visible at first but soon fading from sight as they swim downward with undulating motion. This is a viviparous species. In the Gulf of Maine gravid females, each carrying a pair of young, have been taken in winter." » Kendall, 1914, p. 1S6. 38 BULLETIN OP THE BUREAU OF FISHEBIES Food. — The mackerel shark feeds on small fish, especially on mackerel and no doubt also on herring (which are an important article in the diet of its European congener) as well as on such other schooling fishes as shad and menhaden. It is also known to eat hake and squid. We find no record of its eating Crustacea, nor do fishermen report it as doing so. Commercial importance. — At the present time the mackerel shark is not of any practical value in the Gulf of Maine. On the contrary it is often a serious nuisance from its habit of rolling itself up in an inextricable snarl of t'nane when it entangles itself in drift or gUl nets. Many years ago shark oil was prized by curriers, and the livers of this species were tried out in considerable quantity, but this was never more than a minor industry, abandoned before the middle of the past century. It is interesting to read, however, that as much as 11 gallons of oil have been obtained from the liver of a single shark 9 feet long, and report has it that the richness of the livers in oU fluctuates over periods of years. Fig. 13. — Sharp-nosed mackerel shark (/surus ti^ris) 12. Sharp-nosed mackerel shark {Isurixs tigris Atwood) Jordan and Evermann {Isurus dekayi Gill), 1896-1900, p. 48. Garman, 1913, p. 36. Description. — This shark so closely resembles the common mackerel shark that I need merely point out the points of difference. Most obvious of these is that while in the latter the first dorsal originates above the armpit of the pectoral, in I. tigris it stands altogether behind the inner corner of the latter, and the second dorsal originates a short distance in front of the anal. Its snout, likewise, is sharper, its pectorals narrower, and there is a color difference. Size. — About the same size as the porbeagle; that is, growing to a maximima length of about 10 feet. Color. — Dark bluish-gray or bluish to ashy brown above, white below, and without the black spot on the pectoral fin so characteristic of the common mackerel shark. General range. — Gulf of Maine to the West Indies. Occurrence in the Gulf of Maine. — As I have pointed out above, there is no know- ing how many of the "mackerel sharks" reported by fishermen in the GuK of Maine may actually belong to this and not to the preceding species. However, not only is it nowhere common so far as known, but its center of abundance seems to be FISHES OF THE GULF OF MAINE 39 south of Cape Cod. The only definite Gulf of Maine records for it, so far as we can learn, are as follows: Off Seguin Island, Casco Bay, Provincetown, Cape Cod, and Massachusetts Bay. We have not seen it. It has been netted in Vine- yard Sound as late in the season as December, and occurs as far south as the Gulf of Mexico. On the other hand it is known to wander as far north as Maine. Habits and food. — This shark is a more slender fish than the common mack- erel shark — large, powerful, and swift-swimming, feeding upon small fish and squid. Little is known of its habits, though what has been written of its relative, /. punctatus, probably applies equally to /. iigris. Its breeding habits are not known. Fig. 14. — White shark ( Carcharodon carckarias) 13. White shark (Carcharodon carcharias Linnaeus) Man-eater shark Jordan and Evermann, 1896-1900, p. 50. Garman, 1913, p. 32. Description. — -The white shark is of the general "mackerel shark" appear- ance, with firm lunate tail, the upper lobe only slightly longer than the lower, triangular first dorsal of moderate size originating over the armpits of the pecto- rals, which are sickle shaped, and roughly twice as long as broad. The second dorsal and anal fins are very small, and the root of the tail bears a well-marked keel on either side. The snout is pointed. Unfortunately there is no obvious "field mark" to distinguish a small white shark from the common mackerel shark when seen swimming, for while the former is the slimmer fish the difference in form is not great. Once captured, however, no confusion could arise, for instead of the slim catlike teeth of the porbeagle we find the man-eater best armed of all modern sharks, its teeth large and triangular and similar in shape in the two jaws though broadest in the upper, with nearly straight cutting edges and serrated margins. As a precaution, however, any wry large, active shark, upwards of 18 feet (3 fathoms) long, with the tail not long (out of ordinary proportions) should be looked upon with suspicion — it might prove to be a man-eater. If it were sluggish, resting with the dorsal fin high out of water, it would no doubt be a harmless basking shark (p. 41). 102274— 2ot 4 40 BULLETIN OF THE BUREAU OF FISHERIES Size. — This is one of the largest sharks, groAving, it is said, to a length of 40 feet or even more. In the British Museum there are the jaws of a specimen 36 feet long. In a shark as large as this the teeth are about 3 inches long. A white shark 12 feet S inches long, taken near Woods Hole, was estimated to weigh 1,000 pounds. Color. — -Back slaty or leaden gray, shading gradually to the white of the under parts. In the porbeagle the transition on the sides from dark back to pale belly is more sudden. There is a black spot in the armpit of the pectoral fin, but neigh- boring parts of fin and body are white. Doi'sal, pectoral, and caudal fins are darkest at their rear margins, but the ventrals are darkest (olive) along the forward edge, fading rearward to white. General range. — Cosmopolitan in tropical and warm-temperate seas, straying northward at rare intervals as far as New England and casually to Banquereau Bank off eastern Nova Scotia." It is apparently rare every\vhere. Occurrence in the Gulf of Maine. — The only reliable Gulf of Maine records of this iU-omened shark are of two small ones mentioned by Storer as taken by Massa- chusetts fishermen between 1820 and 1850; one about 13 feet in length and weighing about 1,500 pounds, killed at Provincetown in June, 1848, which he described under the name C. atwoodi; another captured at Eastport, Me., in 1872; one 7 feet 23^ inches long taken many years ago in Massachusetts Bay (figured by Garman, Memoirs, Museum of Comparative Zoology of Harvard University, Vol. XXXVI, 1913, pi. 5, figs. 5-9); and one 16 feet long, taken in a trap at East Brewster, Mass., October 16, 1923, and identified by Doctor Garman. Captain Atwood -* also writes that he saw four caught in mackerel nets at Provincetown. Several more (all rather small) have been taken at Woods Hole in the fish traps, and one off South Amboy, N. J., on July 14, 1916. So seldom does this tropical shark stray to the Gulf of Maine that it would deserve no more than the briefest mention were it not the onh^ shark likely to attack hxunan beings. Being equipped as it is with a most terribly effective set of cutting teeth, and strong and active, the white shark has borne an unsavory reputation as a man-eater from the earliest times, and it was probably a small "man-eater" — in fact, the specimen listed above from South Amboy — that was responsible for the shark fatalities along the New Jersey beach in July, 1916 (p. 22). Hence, so long as white sharks do occasionally wander within om- limits the possi- bility of similar attacks on bathers along beaches of Massachusetts is always open, if exceedingly remote. So far as we can learn, however, there is no actual record of a white shark wantonly attacking human beings in the Gulf (p. 22), but Captain Atwood tells us of a case where a rather small one (apparently the 13-foot specimen described by Storer) turned furiously on a boat but was eventually lanced to death and brought into Provincetown. It is on record, also, that one about 13 feet long attacked a fisherman in a dory on Banquereau Bank many years ago, leaving in the sides of the boat fragments of its teeth, by means of which Doctor Garman was able to identify the species to which the shark belonged. ^° '■ Putnam. Bulletin, Essex Institute, vol. 6, 1S74, p. 72. Salem. '« Quoted, by Goode et al, 1884, p. 671. ''1 Putnam. Bulletin, Essei Institute, vol. 6, 1874, p. 72. FISHES OF THE GULP OF MAINE 41 Habits and food. — So rare (and fortunately so) is this shark even in the tropics that practically nothing is known of its habits. It feeds on large fish, on sea turtles, and perhaps on porpoises. Off the California coast sea lions also faU prey to it — vide Jordan and Evermann's account of a young sea lion of 100 pounds weight in the stomach of a 30-foot white shark. As to its breeding habits nothing is known, though presumably it is viviparous like its close relatives. Fig. 15.— Basking shark (Cctorhinus maiimua) 14. Baskiug shark {CetorMnus maximus Gunner) Bone shark Jordan and Evermann, 1896-1900, p. 51. Garman, 1913, p. 39. Description. — The basking shark resembles the other mackerel sharl^ in its lunate tail, which is much broader than long and with the lower lobe but little shorter than the upper; in the presence of a strong "fore and aft" keel on either side of the root of the tail; in the fact that the second dorsal fin is very much smaller than the first; and in its form, tapering in both directions to snout and tail. However, it is set apart from all other sharks by the enormously long gill slits, which extend nearly right around the neck, and — even more significant — that alone of all its tribe, except its relative the whale shark (Rhinodon), it has rakers on its gill arches, suggesting (though not corresponding to) those of herring, menhaden, etc., among bony fishes. It was the fancied resemblance of these rakers to the whalebone of the whalebone whales that suggested the vernacular name "bone shark" to the whalemen of olden times. Corresponding to its feeding habits, the mouth of the basking shark is very large, but its teeth are very small though numerous. I need only note further that the triangular first dorsal fin stands midway between pectorals and ventrals, and though the back fin is little longer in proportion than that of the other mackerel sharks it rises high in the air when the fish lies awash on the surface, as is its habit — a valuable field mark (p. 39). The nose of large specimens is of ordinary "shark" outline — short, conical, blimtly pointed. In young fish, however, up to 12 or 13 feet in length, it is curiously contracted in front of the mouth into a semicylindrical snout pointed at the tip. 42 BULLETIiSr OF THE BUREAU OF FISHERIES Size. — The basking shark grows to a length of at least 45 feet, perhaps larger. Several specimens 28 to 35 feet in length have been recorded from the New England coast, and stUl larger ones have been reported, but on doubtful evidence. Color. — This shark is grayish-brown, slaty, or nearly black above. The under parts are usually described as white, but the Menemsha specimen recorded by Allen " was of a lighter shade of slate below than above, and one 14 feet long captured at West Hampton, Long Island, on June 29, 1915,-^ had the belly as dark as the back, the only white being a patch underneath the snout in front of the mouth. General range. — This enormous fish is usually said to be native to Arctic seas, straying southward to Portugal on the one side of the Atlantic, to Virginia on the other side, and to California in the North Pacific. It would, we tliink, be more accu- rate to say that it roams the whole North Atlantic from latitude about 35° north to Iceland and northern Norway, Smitt ^ having shown that it is not, strictly speak- ing, an Arctic fish, and that the old tales of a tremendous whale-eating shark, on which Fabricius based his statement that the basking shark occurs in Greenland seas, were false. It is also plentiful enough off the coasts of Ecuador and Peru in the South Pacific to support a considerable local fishery." Occurrence in the Gulf of Maine. — Of recent years the bone shark has been seen but seldom in the Gulf of Maine, the list being as follows: One 28 feet long was killed in Maine waters in 1828; one off Musquash Harbor in the Bay of Fundy in August, 1851; one of 34 feet at Eastport in 1839; several ranging in length from 25 to 35 feet, killed there in 1868 and 1870; a considerable number seen and several secured off Cape Elizabeth in 1848 by a whaler cruising for humpback whales; one of 35 to 38 feet harpooned but lost between Boston and Provincetown in 1864; and one killed near Provincetown in 1835, another in 1836 or 1837, a third in 1839, and a fourth in 1847. We do not find another definite record of the bone shark in the Guff of Maine until October 8, 1908, when one 18 feet long (measured by J. Henry Blake) was taken in a weir near Provincetown. Two more have been killed there since — one a 22-foot fish on October 9, 1909, and the other of 29 feet on June 8, 1913, both in the harbor. Mr. Blake also reported one of 31 feet (16 feet in girth) as taken at Long Point, near by, but the year is not recorded. A small one of 12 to 14 feet was caught at Menemsha Bight on Marthas Vineyard on August 16, 1916, and one of about 26 feet 6 inches * at the same locality on June 24, 1920. The bone shark is so large a fish and so conspicuous, thanks to its basking on the surface, that every specimen visiting the coastwise waters of the Guff is almost certain to be seen sooner or later and to be harpooned. Hence it is probably no commoner there than the meager record suggests. >» Bulletin, Boston Society of Natural History, No. 24, March, 1921, p. 5. SI This specimen is described by Hussalcof (Copeia, Aug. 24, 1915, No. 21, pp. 25-27). « Scandinarian Fishes, 1892, p. U46. " This fishery is described by Stevenscn (Report, U. S. Commissioner of Fisheries for 1902 (1904), p. 228). ^' This specimen is now preserved, mounted, in the Boston Society of Natural History, and described by Allen (Bulletin^ Boston Society of Natural History, No. 24, March, 1921, pp. 3-10), who collected the foregoing records. riSHES OF THE GULF OF MAINE 43 Before the coming of the white man this great shark seems to have been a I egnlar inhabitant of the Gulf of Maine, which afforded it an excellent pasture, for old tradition has it that large numbers were taken in Massachusetts waters for their oil during the first half of the eighteenth century. However, the local stock soon went the same way as the local stock of the North Atlantic right whale — into the try pot — and this seems also to have been its fate in Norwegian waters, where it was sufficiently abundant to support a regular fishery up until about 1820, since which time it has been killed down to but a fraction of its former numbers. Indeed, the basking shark to-day is something of a rarity off the coast of Norway, but in other parts of the world, particulai'ly in Icelandic waters, off Ireland, and off Peru, as noted elsewhere (p. 42), it is still moderately plentiful. Habits. — This is a sluggish, perfectly inoffensive fish, helpless of attack so far as its minute teeth are concerned, and spending much time sunning itself on the surface of the water, often lying with its back awash, on its side, or even on its back, and sometimes loafing along with the snout out of water. Hardly a writer men- tioning this shark but tells us that two or three swimming tandem, with the dorsal fins high in the air, are the basis for "sea-serpent" myths. At times bone sharks are gregarious, traveling together in schools. Nothing whatever is known of the breeding habits of the basking shark. Food. — Next to its vast bulk and its curiously sluggish habit, the most inter- esting peculiarity of the basking shark is its diet, for it subsists whoUy on minute Crustacea, particularly on copepods, and on other tiny pelagic animals, which it sifts out of the water by means of its greatly developed gill rakers, exactly as do such plankton feeders as menhaden on the one hand and whalebone whales with their baleen sieves on the other. Commercial importance. — Although the day of the bone shark in New England waters is long past, probably never to return, it may be of interest to point out that it has always been hunted whenever encountered by the sperm whalers from New Bedford, and that it is still an object of pursuit off the coasts of Iceland and Ireland. It was and is valued solely for its liver oil, individual fish as a rule yielding from 80 to 200 gallons (average about 125 gallons), with as much as 400 gallons from a single liver not unheard of and a yield of GOO gallons reported. The basking-shark fishery has always been carried on with harpoons, the shark being quite indifferent to the approach of a boat though it swims actively and strongly when struck. Fat ones are subdued more easily than lean ones. 44 BULLETIN OF THE BUREAU OF FISHERIES THE SPINY DOGFISHES. FAMILY SQUALID.^ This group is characterized and made easily recognizable by the presence of two dorsal fins, each with a fixed spine, but no anal fin, while the teeth are alike in the two jaws in some, unlike in others. 15. Spiny dogfish {Squalus acanthias Linnseus) Dogfish; Piked dogfish; Gkayfish Jordan and Evermann, 1896-1900, p. 54. Garman, 1913, p. 192. Description. — So rare are all other spiny sharks in the Gulf of Maine that any little shark with a large sharp spine close in front of each dorsal fin caught there is practically sure to be a "dog," of which there are thousands in the Gulf to one of any other shark. Should the fish be uniform dark brown or black it might possibly (but not probably) prove to be the black dogfish (p. 52). A glance at the tail fin will settle the question, for the rear margin of the latter is deeply notched near its tip (fig. 19), whereas in the common spiny dog its margin is entire. This is a slender little shark with tapering but rounded head and flattened snout. Its first dorsal fin stands between pectoral and ventral; its second dorsal is about Fig. 16.— Spiny dogfish (Sgualus acanthias). After Oarman two-thirds as large as the first; its pectoral is triangular, broader at the base than it is long; the lower lobe of the tail fin is well marked; and the ventrals are well forward of the second dorsal. The spines are close up against the front margins of the two dorsals, the first shorter and the second nearly as long as their respective fins are high, and they are very sharp, as every fisherman knows to his cost. The spiny dog has no anal fin, a lack separating it from all smooth-finned sharks known from the Gulf of Maine, except the Greenland shark (p. 53). There is a low fold of skin on either side of the root of the tail back of the second dorsal fin, so small, however, that there is no danger of confusing it with the keels of the mackerel sharks. The teeth are small, their sharp points bent toward the outer corners of the mouth and each row forming a continuous cutting edge. Size. — Mature dogs are ordinarily 2 to 3J4 feet long. Mature males grow to a length of about 3 feet and a weight of 5 to 6 pounds; females to 3 or 3 }4 feet and a weight of 8 pounds. Occasionally very large fat specimens may reach a weight of 15 pounds. FISHES OF THE GULF OF MAINE 45 Color. — Usually slate colored above but sometimes brown, with a row of small white spots on each side from the pectoral to abreast of the anal, and a few other white spots in front of and behind the first dorsal and in front of the second dorsal fins. These spots are most conspicuous in small fish up to 12 or 14 inches in length and fade with growth until in some specimens they disappear altogether. It is gray to white below. General range. — Both sides of the North Atlantic, also Mediterranean; on the American coast from the Gulf of St. Lawrence and the banks of Newfoundland south to Cuba. Replaced by closely allied species in the North and South Pacific and Indian Oceans. Occurrence in the Gulf of Maine. — The spiny dogfish — "dogfish" or "dog" in common parlance — makes up for the comparative rarity of other sharks in the Gulf of Maine by its obnoxious abundance. To mention all the localities from which it has been reported there would be simply to list every seaside village and every fishing ground from Cape Cod to Cape Sable. On the offshore banks, too, it is as familiar as it is along the coast. Dogfish are seasonal visitors. In spring they strike in almost simultaneously along the whole coast from New England to North Carolina, appearing at Cape Lookout in April, off Long Island abundantly in May, and as early in the season on Georges Bank (April-May) as at Cape Lookout. In the inner parts of the Gulf of Maine the date of the first heavy run of dogfish varies widely from year to year and from place to place. We have not heard of them in Massachusetts Bay before May. Indeed, summer warming is hardly appreciable more than a few fathoms below the surface until well into that month, so they could hardly be expected earlier. However, according to reports of local fisher- men the period of freedom may close there as early as the last half of the month in some years. In 1903, for example, they appeared as far north as Penobscot Bay by the middle of May, and though as a rule it is not until June that they arrive in numbers in the Massachusetts Bay region, it is sometimes impossible to set gill or drift nets anywhere between Cape Cod and Cape Elizabeth after the first days of that month, so numerous are they. In 1913 the first heavy run of dogs struck Ipswich Bay on June 14, and they appeared there at about the same date in 1905, but there is much local variation in this respect. In 1903, for example, they did not appear until early July at Provincetown, though swarming a month earlier in other parts of Massachusetts Bay, in Ipswich Bay, and off Penobscot Bay. However, they usually strike in all along the northern Maine and west Nova Scotian coasts by the end of June, though earlier in the open Bay of Fundy than in Passamaquoddy Bay, where few are seen until late in July. West of Cape Cod (that is, at Woods Hole and oft" Long Island) it was formerly believed that these little sharks were only transients, passing north in spring, south in autumn, which were the only seasons when they were seen inshore regularly. However, dogs, both large and small, are caught in the traps of the Woods Hole region in Julj^, and Latham's ^^ recent discovery that adult spiny dogfish are common in deep water in Long Island Sound in summer, together with the fact (on which he comments) that young ones are taken in great numbers in the traps on Long Island « Copeia, Oct. 15, 1921, No. 99, p. 72. 46 BULLETIN OF THE BUREAU OF FISHERIES in July and August, is sufficient proof that while some of the fish that visit the middle Atlantic coast in spring may go north to the Gulf of Maine, others merely drop down into deeper water to summer, coming inshore again for a time in autumn. Most of the dogfish take their departure from the inner parts of the Gulf during October, few being caught on the coast north of Massachusetts Bay after November 1. Rarely, however, they stay later, as in 1903 (a big dogfish year), when they were abundant along the outer shore of Cape Cod as late as the third of the month. Ordinarily none are caught within the Gulf of Maine north of Georges Bank in ^\^nter, but this, like most rules, has its exceptions. In 1882, for example, schools 6000 7500 - 7000 - 6500 - 6000 2 5500 I— £ 5000 g 450 3= 4000 g 3500 3000 2500 2000 1500 1000 SCO ii. J_ _L_l i. MAR. APML MAY JUNE JULY AUG. SEPT. OCT.' NOV. DEC. Fig. 17.— Numbers of spiny dogfish caught on certain otter-trawling trips to Georges Bank dtuing the different months of 1913 were reported off Portsmouth in February, while in 1913 a few were caught 20 miles off Cape Ann on November 19 to 24, many near Boon Island from December 5 to 13, and on Jeffrey's Ledge on December 11 and 12. Dogfish appear earlier in spring and linger later into the winter on Georges Bank (fig. 17) than in the inner parts of the Gulf. It is safe to say that there are few there in March, the earliest definite record (obtained during the investigations of 1913) being of 25 fish caught on the "winter cod ground" east of the shoals (longitude about 67°, latitude about 41° 40') between the 20th and the 22d, and of 46 from the same general region from the 27th to the 30th. Their numbers FISHES OF THE GULF OF MAINE 47 increase rapidly in April on Georges Bank, however, and they ai( a pest there after the 1st of May, while the last half of June, July, and August sees them at the height of abundance at least on the portions of the bank resorted to by the otter trawlers at that season. Whether by chance or as reflecting an actual diminution in the stock of dogfish present, the catches fell off markedly during September in 1913, but considerable numbers were taken throughout that month. Very few were actually captured by the trawlers in October, but there was a considerable increase in November, probably reflecting the southward passage of the schools that had spent the summer further east. A few were caught in November and December, and one on the southern part of the bank (latitude about 41°, longitude about 67° 30') as late as January 20 to 22. Thus February is the only month when the bank is entirely free of them. The time table just outlined for the year 1913 may be taken as typical, for it corroborates the various reports of fishermen tabulated by the Massachusetts Commissioners of Fisheries and Game in 1905. Apparently dogfish reach Browns Bank rather later than they do Georges, for none were taken there on April 14 in 1913, though they are only too plentiful there in summer. It is also very likely that they depart thence rather earlier, though a few lingered as late as December 3 to 12 on Western Bank off Halifax in that year. Gravid females have been described as arriving before the males in spring, but this remains to be confirmed. The accompanying graph (fig. 17) of the nximbers of dogs taken by certain otter trawlers on Georges Bank at various dates during the year 1913 will more graphically illustrate the seasonal fluctuations of this fish there, with the reservation that the precise catches are governed not only by the abundance of the stock but also by the precise grounds fished on and by the general success of the sets. The winter home of the Gulf of Maine dogfish is still to be learned. They have often been said to migrate south to the Tropics, and it is certain that some dogfish do reach Cuba during the cold season, but the fact that they appear so nearly simul- taneously all along the coast north of North Carolina in spring, and that they leave Georges Bank so late in the season, with the discovery of dogfish in deep water in Long Island Sound in siammer (p. 45) argues for an on-and-off rather than a long- shore migration, with the deep water off the continental slope as their winter home. This is corroborated by the fact that on February 20 to 21, 1920, the Albatross trawled several specimens in depths of 90 and 199 fathoms along the continental edge off Chincoteague, Va., and off Delaware Bay. Also, they are usually so thin when they appear in spring that they can feed but little during the winter. In short, evidence is gradually accimiidating to the effect that the seasonal movements of the spiny dogfish parallel those of the mackerel (p. 191). It is generally beheved that dogfish not only summer more regularly in the region of Massachusetts Bay now than of old, but that they are far more numer- ous there than during the first hah of the past century. At Woods Hole, on the contrarjT, they and the smooth dogfish were much more plentiful before 1887 than at any time since then. To a certain extent, of course, reports of fluctuations in abundance from year to year must be discounted as reflecting the movements 48 BULLETIN OF THE BUREAU OF FISHERIES of the great schools that may visit one part of the coast one summer and another part the next, there being no general alteration of the stock, but the many fishermen who reported to the Massachusetts Commissioners in 1905 were so unanimously of the opinion that dogfish had multiplied steadily for 20 to 30 years past as to point unmistakably to the conclusion that the species as a whole was then in one of the periodic upswings characteristic of various other fishes. Reports from British coasts are to the same effect. Perhaps the years 1904-5 marked the apex of this wave of multiplication; at any rate dogfish were reported as distinctly less trouble- some to the mackerel netters in 1913 than in previous years, and since that time less complaint has been made of them, though it is too soon to say whether a general diminution of the stock is actually in progress. Much has been written of the habits of the spiny dogfish, all to the effect that it has nothing to recommend it from the standpoint either of the fishermen or of its fellow creatures in the sea. It is one of the more gregarious of our fishes, swi mm ing in schools or packs. Swedish fishermen assert that young dogs school separately from their parents, and it is certain that fish of a size continue to associate together as they grow, the result being that any given school runs very even, consisting as a rule either of the very large mature females, of medimn-sized fish (either mature males or immature females), or of small immature fish of both sexes in about equal numbers. '' Apart from its general seasonal migratory movements, the dogfish are governed by the movements of the fishes on which they prey and in pursuit of which they roam about, striking in here and there in multitudes. Fortunately they seldom stay long in one place, but there is seldom, if ever, a time during the summer when they are not common on some part of the Gulf of Maine coast. So erratic are their appearances and disappearances that where one has good fishing to-day he may catch only dogfish to-morrow and nothing at all the day after, the better fish having fled these sea wolves and the latter departing in pursuit. The dogfish use their back spines for defense, curling around in a bow and striking, which makes them hard to handle on the hook. It is probable, too, that the spines are slightly poisonous, general report to this effect being corroborated by the fact that the concave surfaces are lined with a glandular tissue resembling the poison glands of the venomous "weever" {Trachinus draco)?'' Strong, swift-swimming, voracious almost beyond belief, the dogfish entirely deserves its bad reputation. Not only does it harry and drive off mackerel, herring, and even fish as large as cod and haddock, but it destroys vast numbers of them. Again and again fishermen have described the sight of packs of dogs dashing among schools of mackerel, and even attacking them within the seines, biting through the net, ruining the gear, and releasing such of the catch as escapes them. Often, too, they bite groundfish from the hooks of long lines, take the baits and make it vain to fish where they abound. In Massachusetts and Ipswich Bays, 3« Ford (Journal of the Marine Biological Association of the United Kingdom, new series, Vol. XH, No. 3, Sept., 1921, pp. 468-505, Plymouth, England) has recently published very interesting notes on this and other phases of the life-history of the spiny dogfish, with a summary of the earlier statements as to the breeding season. " Dale (Philosophical Transactions, Royal Society of London, series B, Vol, 212, 1923, p. 27) describes the spines and gives clinical records of the effects of wounds inflicted by them. FISHES OF THE GULF OF MAINE 49 indeed, as well as about the Isles of Shoals, hook-and-line fishing is often actually prevented during the period of summer plenty unless cockles be used for bait, for dogfish do not take these. When schools of dogfish rush headlong into net or seine, as often happens, they so snarl the twines that disentanglement and re- pair may be the work of days, and it has been estimated that they do no less than $400,000 worth of damage annually to fishing gear and to fish caught by such gear off the Massachusetts coast alone — probably no less along the shores of Maine, so that in the aggregate they are a heavy debit in the economic scale. Rumor has it, even, that packs of dogfish have been known to attack swimmers and liter- ally bite them to pieces, but we can not vouch for this. At one time or another they prey on practically all species of Gulf of Maine fish smaller than themselves, and squid are also a regular article of diet whenever they are found. Dogfish are also known to take worms, shrimps, prawns, and crabs, and when they first arrive at Woods Hole from the south in May they are often found full of Ctenophores, being one of the few fish that eat these watery organisms. It would be pure guesswork to attempt to estimate the actual numerical strength of the dogfish, but they must be plentiful, indeed, when they can often be caught as fast as they can be hauled in, when line trawls with 1,500 hooks have brought in a dogfish on nearly every hook, and when as many as 20,000 have been recorded in a single draught of a seine in British waters. Breeding habits. — -From time immemorial fishermen have known that the spiny dogfish is viviparous. Aristotle, indeed, describes its manner of bearing young. The eggs are large, well stored with yolk, and during early stages of develop- ment those in each oviduct (the so-called "uterus") are contained in a horny capsule that later breaks down, leaving the embryos lying free in the "uterus" with which they have no placental attachment. Ford's studies, men- tioned above, suggest about 10 to 11 months as the period from fertilization to birth, which takes place when the young are 9 to 12 inches (23 to 31 cm.) long, and as they are then practically of adult form with the yolk almost wholly ab- sorbed, strong and active, their chance of survival is excellent. Ordinarily a female has 3 or 4 young to a litter — sometimes as few as 1 or as many as 8 to 11 — and while the embryos are developing in the uteri a fresh set of ovarian eggs is growing, ready to take their place. It has often been suggested that the dogfish may give birth to 2 or 3 litters — -that is, upwards of 20 pups — annually, but if Ford's estimate of the duration of gestation is correct one litter per year would be the rule. State- ments as to the season at which the young are born are conflicting. At Plymouth, England, this takes place from January until March, according to Garstang; from August until December, according to Ford. This, of course, suggests two dis- tinct breeding seasons, and we believe that, similarly, among the dogfish that visit the Gulf of Maine some females give birth to their young in late autumn, others in late winter or early spring. For the evidence on which we base this view we are indebted to Dr. H. V. Neal, whose acquaintance with dogfish on the Maine coast is very intimate. It has long been known that when the dogs first appear on the Massachusetts coast in May or June many of the females contain embryos 50 BULLETIN OF THE BUREAU OF FISHERIES of some size, which, as Doctor Neal tells us, grow to 4 to 7 inches in length by July. However, during this same month other females caught along the coast of Maine are found to contain embryos in very early stages of development, from the formation of the germ ring to a length of about 4 mm. By September the embryos of the older generation range from 7 to 11 inches in length, some of them being almost ready to be born, while those of the younger generation (any given female contains only embryos of one or of the other generation, never of both) have grown to an average of about 17 mm. Probably the older generation is born in October and November, while the younger one winters in the uterus of the mother, to be born in spring. Fall-bearing females are then fertilized again, the development of the next set of eggs commencing in the early winter, while spring-bearers are fertilized in early summer, which corroborates 11 months as the known period of gestation (p. 49). This would also explain the fact that dogfish smaller than a foot in length are never reported in the Gulf of Maine, for the young are produced during the season when there are very few dogs on this coast, these few probably being immature. In short, the inner parts of the Gulf of Maine probably do not serve as a nursery for the dogfish, plentiful though this fish is there in summer, but the young are born somewhere offshore and probably while the parents are in deep water. It seems, however, that this seasonal schedule does not apply west of Cape Cod, for Latham^* records a great abundance of very young ones taken in the traps in Long Island Sound in August, showing that one generation is produced there in midsummer. Dogfish only 1 foot long, hence new born, have been found in the stomach of a goosefish at Woods Hole in July (p. 527). Commercial value. — With the dogfish so destructive to fish and to gear, and with so many of them caught both by lines and by otter trawls during more than half the year, it is no wonder that serious efforts have been made to utilize them on a large scale — to make them marketable and a source of revenue instead of a dead loss. Since this matter has been the subject of discussion elsewhere we need point out only that the dog is a far better food fish when fresh than is generally appre- ciated, and that it would offer a tremendous supply of cheap food were a satisfac- tory method of canning it to be worked out. Dogfish have also been used in the manufacture of fertilizer, and enough dogfish livers are brought into New England fishing ports to yield almost 10,000 gallons of oil annually, which is combined and sold with cod-liver oil. Up to the present, however, dogfish have not been of sufhcient value to compensate for a hundredth part of the damage they do and most of those caught are thrown back into the sea.^' '8 Copeia, Oct. 16, 1921, No. 99, p. 72. 3" For further discussion of the damage done by dogfish and of their commercial possibilities, see the following: " Report upon the damage done by dogfish in the fisheries of Massachusetts," Annual Report, Commissioners of Fisheries and Game [of Massachusetts] for 1905 (1906), pp. 97-169; "Aquatic products in arts and industries," by Charles H. Stevenson. Report of the Commissioner, U. S. Commission of Fish and Fisheries, Part XXVIII, 1902 (1904), pp. 228-229; Field, 1907, pp. 12-18, 40-19; "Sea mussels and dogfish as food," by Irving A. Field. Proceedings of the Fourth International Fishery Congress. In Bulletin. U. S. Bureau of Fisheries, Vol. XXVIII. 1908 (1910), pp. 243-257; and Mavor, 1921, pp. 125-135. FISHES OF THE GULF OF MAINE 51 16. Portuguese shark (Centroscymnvs cmlolepis Bocage and Capello) Jordan and Evermann, 1896-1900, p. 55. Carman, 1913, p. 204. Description. — This shark can easily be identified by the fact that while its general appearance — particularly the absence of anal fin, the situation of the ven- trals far back under the second dorsal, and its rather stout form and blunt snout — might lead a hasty observer to think he had caught a small Greenland shark, more careful examination, by touch if not by eye, would reveal a short spine close in front of each dorsal fin. The first dorsal is smaller than in any of our sharks except the "Greenland," the second is about as high as the first, and the ventrals are larger than either. The tail is notably short and broad and its upper lobe is notched. The teeth are very different in the two jaws — narrow, pointed, and of the seizing type in the upper; broader, oblong, with a notch on one side near the tip, and forming a cutting edge in the lower. Sise. — -Adults run from 3 to 4 feet long, as they are caught. Garman records one 44 inches long off the coast of New England, but 10 inches is the smallest we find mentioned. FiQ. 18. — Portuguese shark {Cevtroscymnus ccelolepis). After Garman Color. — Described as deep chestnut brown on the belly, as well as the back. General range. — This rare deep-water shark, originally known from off Portugal, has since been taken at various other localities.^" Its claim to mention here rests on the fact that it was once reported off Gloucester; on the specimen "taken off the coast of New England," just mentioned; and on Goode and Bean's (1896) statement that it is abundant on the slopes of our offshore banks at 200 fathoms and more. Habits. — Little is known of its habits beyond the fact that it is a deep-water species regularly caught by Portuguese fishermen with hand lines, a fishery that Wright (Annals and Magazine of Natural History, series 4, Vol. II, 1868, p. 426) describes as follows: Some 600 fathoms of rope were let out, the first 30 or 40 fathoms of which had fastened to it at intervals of a fathom a series of small ropes, on each of which was a large hook baited with a codling. This fishing tackle remained below for about two hours, when they commenced to haul it in. When it arrived at the last few fathoms, they pulled in, one after another, five or six specimens from 3 to 4 feet long. The species was the Centroscymniis ccelolepis Bocage and Ca- pello. These sharks, as they were hauled into the boat, fell down into it like so many dead pigs. This species is viviparous, 13 to 16 young having been found in females caught off Portugal. " Known from Portugal, the Mediterranean, Madeira, Japan, the Faroes, and recently reported from Iceland by Ssemunds- son (Videnskabelige Meddelelser fra Dansk naturhistorisk Forening i Kj0benhaven [Copenhagen], Bind 74, 1922, p. 167). 52 BULLETIN OF THE BUREAU OF FISHERIES 17. Black dogfish {CentroscyUium fdbricii Reinhardt) Jordan and Evermann, 1S96-1900, p. 56. Garman, 1913, p. 231. Description. — As pointed out elsewhere (p. 44), the notched margin of the upper tail lobe distinguishes this rare shark at a glance from the spiny dogfish, with which it agrees in the possession of a long pointed spine at the front edge of each dorsal fin, the second being longer than the first. It differs further in that its dorsal spines are deeply grooved on each side, whereas in the "dog" they are rounded; in the location of the ventral fins, the rear axils of which stand almost directly under the front origin of the second dorsal instead of some distance in front of it; in its small pectorals of rounded outhne; in the structure of its teeth, each of which is tridentate, with sharp points ; in its broad rounded snout ; and in its very dark color. Like the spiny dogfish, it lacks an anal fin. Sise. — The specimens so far described have ranged from 2}4 to 3J^ feet in length — that is, about the same size as the spiny dogfish. Color. — Uniform dark bro'wn to black, below as well as above. General range. — Positive records for this shark are from Greenland, Iceland," rather deep water off the outer banks, Grand to Georges,^ off the Hebrides and Faroes where two specimens were taken by the Norwegian fisheries steamer Michael Sars in 400 to 600 fathoms, and from the North Atlantic (two specimens in the British Museum) . But since Tate-Regan *^ thinks a specimen that he examined Fig. 19. — Black dogfish ( CentroscvlUum fabricii). After Garman from the Falkland Islands is identical, while Goode and Bean (1896) tentatively refer to it a young shark from the Gulf of Mexico, and the Japanese C. ritteri seems hardly distinguishable, the black dogfish may prove to have a cosmopolitan range in deep waters. Occurrence in the Gulf of J/aine.— Evidently the black dogfish is very rare in the Gulf of Maine, for it has so far been reported there only from Georges Bank, from the slope off Browns in 200 fathoms, and vaguely from off Gloucester, which might mean any of the fishing grounds between Cape Cod and Newfoundland. However, it has been taken repeatedly on the offshore slopes of the Nova Scotian Banks in 200 to 250 fathoms, whence a number were brought into the Bureau of Fisheries by halibut fishermen many years ago." Nothing is known of its habits. " Sffimundsson. Videnskabelige Meddelelser fra Dansk naturhistorisk Forening i Kj0benhaven, Bind 74, 1922, pp. 159-205. " According to Garman (I9I3), Greenland to New York. " Aimals and Magazine of Natural History, Vol. II, Eighth Series, 1908, p. 49. London. " For list of these specimens see Bean (1881, p. 116). FISHES OP THE GULF OF MAINE 53 THE NURSE SHARKS. FAMILY SCYMNORHINID^ The nurse sharks, like the spiny dogfishes, lack anal fins, but there are no spines in their dorsal fins and the teeth in the upper jaw are noticeably unlike those in the lower. IS. Greenland shark {Somniosiis microcephalus Bloch and Schneider) Nurse shark; Sleeper shark; Gurry shark; Ground shark Jordan and Evermann, 1896-1900, p. 57. Garman, 1913, p. 241. Description. — The Greenland shark is notable for its very small dorsal fins, without spines, the second being of about the same size as the first, and for small pectorals hardly larger than the ventrals, coupled with the absence of an anal fin and with a tail of more "fishlike" form than that of most other sharks except the mackerel-shark tribe. Bearing these points in mind, particularly the absence of anal fin and dorsal spines, it can not be confused with any shark common in our Gulf. The location of the first dorsal — about midway between pectorals and ventrals — is the most obvious "field mark" to distinguish it from the rare Echino- rhinus hrucus (p. 55). We may note further that the Greenland shark is compara- FiG. 20. — Qreenland sbark iSomniosus Tnicroccphatus). After Garman tively stout shouldered, tapering thence toward the tail; that its snout is blunt and rounded as Scoresby *^ represented it a century ago (many more recent figures of it are caricatures in this respect) ; that the gill openings are short and located low down on the sides of the neck; and that the teeth are unlike in the two jaws, being narrow in the upper, and broad, square tipped, and notched at the outer corners in the lower jaw. Size. — This is one of the larger sharks. It is said to grow to a maximum length of 24 feet, but few, if any, actually reach such a size, 18 feet being unusual. One 15 feet long has been taken in Cape Cod Bay; another of 133^ feet (now in the Museum of Comparative Zoology) in Massachusetts Bay. Perhaps S to 12 feet would be a fair average for adults; nor is this size exceeded often among the hundreds annually caught about Iceland and Greenland. General range. — Arctic seas; south to Cape Cod in the western North Atlantic, and to France in the eastern North Atlantic; to Oregon in the Pacific. It is the object of a regular fishery in Greenland, Iceland, Norway, and Spitzbergen. Occurrence in the Gulf of Maine. — Although there is no reason to suppose that the Greenland shark is ever common in our Gulf or appears there other than as a " An account of the Arctic Regions, and of the whale fishery, 1820, Vol. II, PI. XV, flgs. 3 and 4. 54 BULLETIN OF THE BUREAU OF FISHEKIES straggler from the north, its presence has been signahzed on several occasions. Two specimens, for example, were taken in the neighborhood of St. Andrews in 1915 (one caught in a weir and the other on a long line). It has also been reported off Eastport, SO miles off Cape Elizabeth, near Cape Ann, off Marblehead and Nahant, in Massachusetts Bay, off Barnstable in Cape Cod Bay (where R. E. Smith killed the fish noted above many years ago), at Provinceto'^vTi, and in Cape Cod Bay off the entrance to the Cape Cod Canal, where a large one between 10 and 11 feet long was taken by a trawler in April, 1924. Although the localities of capture are so ^\^dely scattered, the total number of specimens definitely recorded from the Gulf of Maine is not over a dozen. Of recent years this has certainly been so rare a shark within the hmits of the Gulf of Maine that one might fish a lifetime without seeing it, but in old days, when right whales were still plentiful and many of them were killed off the Massachusetts coast, it may well have been more abundant — such, indeed, is the rumor — for in its northern home it is attracted from afar to feed on whale, seal, and narwhal carcasses, from which it gets one of its popular names. When there has been a big killing of narwhals, such as falls to the lot of the Eskimo of Disko Bay at rare intervals, schools of these great carrion eaters may linger in the vicinity for several years. Food. — This is one of the most sluggish of sharks, offering no resistance whatever when hooked, entirely inoffensive " but extremely rapacious, biting on anything in the way of meat, the more putrid and ill-smelhng the better. Apart from carrion, which can be available only at rare intervals, it feeds on fish and seals. Cod, hng, and halibut have been found in its stomach, and an entire reindeer has been found in one. The specimen from Cape Cod Bay, mentioned above, contained half a dozen flounders and a large piece bitten out of the side of a seal. It is also known to eat crabs. An old story has it that the Greenland shark attacks live whales, but this is not confirmed by recent observation and is most improbable. Although so sluggish, apparently it is able to catch live seals, for not only have whole ones been found in its stomach, but when sharks gather seals soon become very scarce. Ealits. — The nurse is a bottom swimmer, seldom coming to the surface except in pursuit of the scent of carrion, such as of a whale being cut up. In Icelandic waters it comes up into water as shoal as 40 to 50 fathoms in winter, but in sunmier descends to 200 or 300 fathoms, lying chiefly on the muddy or clay bottom of troughs or folds in the sea bottom. In the Gulf of Maine, then, it would be more apt to be found in the deep basin than near land. Breeding Jiahits. — Nothing definite is known of its breeding habits. Its close relative, Somniosus hrevipinna, of the Mediterranean, the coasts of Portugal, and of Japan, has long been known to be viviparous, and the early belief was that this also applies to the Greenland shark, Faber stating that its young are born in July and August. However, no one has recently reported a fetus in a Greenland shark, and the fact that females often contain great numbers of eggs (up to the size of ,« Tales to the effect that it attacks GreenJanders in their kyaks are apparently mythical, and Doctor Porsild, director of the biological station at Disko, said that the Eskimos do not fear it as they do the killer whale; nor is there any authentic instance on record of a shark attacking a human being about Iceland. FISHES OF THE GULF OF MAINE 55 goose e^s) all at about one stage of development,*' has led to the common belief in Iceland that it is oviparous. For ourselves, it seems so unlikely that, of two sharks closely alUed in every way, one should retain the fetuses in the oviducts until they are of considerable size and fully developed, and the other lay eggs, that we incline to the belief that the Greenland shark ■ftdll also prove to be viviparous. Commercial importance. — Were the nurse more plentiful in our waters it might be a useful scavenger. Off Iceland it is caught for its liver oil, and in the Arctic the flesh is dried for dog food. It is very interesting to note in passing that while the meat is perfectly wholesome when dried, it produces a sort of intoxicant poison- ing when eaten fresh.*' THE BRAMBLE SHARKS. FAMILY ECHINORHINID^ The only living representative of this family (it is represented among the tertiary sharks) resembles the nurse-shark family (p. 53) in lacking both anal fin and dorsal spines, but its teeth aie alike in the two jaws. 19. Bramble shark (EcMnorhinus hrucus Bonnaterre) Jordan and Evermann {E. spinosus), 1896-1900, p. 58. Garman, 1913, p. 243. Description. — As pointed out above, the location of the first dorsal fin, above the ventrals instead of about midway between the latter and the pectorals, is Fki. 21.— Bramble shark (Echinorhinui irucus) the readiest field mark to enable separation of this form from the Greenland shark. Brucus also differs from the latter in its more slender form, longer gill slits, and especially in the fact that the teeth are alike instead of unlike in the two jaws. Size. — The largest (a specimen from British waters) of which we have found a record was 9 feet long, and it has been credited with a weight of 400 pounds. " Smitt (Scandinavian Fishes, 1892) describes one with "innumerable" small eggs and discusses this question, and Helbing (Nova .\cta, Kaiserlichen Leop.-Carol. Deutschen Akademie der Naturforscher, vol 82, 1904) has recently given a good description and figures of fetuses of Somniosus brevipinna (as Lsemargus rostratiia), with a discussion of the relationship of this species to the Greenland shark (as L^margus borealis). « This is described by Jensen in " The Selachians of Greenland." Saertryk af Mindeskrift for Jepetus Steenstrup, pp. 12-14, 1914. Translation by A. H. Clark, Science, New Series, Vol. XLI, Jan.-June, 1915, p. 796. 56 BULLETIN OF THE BUREAU OF FISHERIES Color. — Described as dark brown above, with or without darker blotches; lower surface lighter to white. General range. — Formerly thought to be confined to the eastern Atlantic, off the coasts of Europe and north Africa, and to the Mediterranean in rather deep water, this shark has since been recorded from the Cape of Good Hope, the Pacific, and from Australia. Apparently, however, it is rare everywhere, unless it be that the rarity of capture is due to its habit of living at considerable depths. Occurrence in the Gulf of Maine. — A single specimen of this little known shark came ashore at Provincetown in December, 1878, and this stiU remains the only record of it from the western Atlantic. Habits. — Nothing is definitely known of its habits or whether it is a regular inhabitant of the continental slope at and below 200 fathoms, as its wide distribu- tion and proclivity for deep water suggests. SKATES AND RAYS Skates, with their disklike outlines, thin as a shingle, and their long tails, are familiar objects along our shores. The Gulf of Maine supports four species in great abundance, while several others have been recorded on rare occasions. So far as the local fauna is concerned, this tribe falls into three groups — first, the skates (family Rajidfe) with comparatively short tails and without spines; second, the sting rays (families Dasybatidse and Myliobatidse) with long whipHke tails armed with stiff spines; and third, the torpedo (family Narcaciontidae) , interesting because provided with electric organs capable of giving a strong shock. All our common species belong to the first group. Among skates and rays, as among sharks, fertilization is internal and the modification of the posterior edges of the ventral fins into rodlike semitubular claspers — the copulatory organs — distinguishes males from females at a glance. Some families are viviparous; others lay eggs. The common skates look so much ahke that fishermen seldom discriminate between them but speak of them all, large and small, simply as "skates." For this reason we know very little about the individual differences in habits between the several species. All, however, live chiefly on or close to the bottom, moving through the water by undulations of the flexible pectoral fins, steering themselves with the tail. All are decidedly omnivorous, feeding largely on the larger Crustacea — shrimps, crabs, lobsters — as well as on mollusks, worms, etc., and to a greater or less extent on fish. In the Gulf of Maine they are a nuisance, for they bite the hook readily and often are caught in great numbers in otter trawls. To give some idea of their abundance on the offshore banks I may note that the average number of skates (all species together) taken on Georges Bank, per trip of 4 to 7 days, on 25 trips by several trawlers, January to December, 1913, was approximately 800, the largest catch being 4,521 skates, the poorest 82. Whether they are equally abundant on Browns Bank is not clear, for though they are famiUar enough there, no statistics as to the actual numbers caught are available. Skates are as plentiful inshore as on the banks, as appears from the following representative catches on long lines: FISHES OF THE GULF OF MAINE 57 1. 13 miles from Gloucester, 2,540 baited hooks. Total fish caught, 540; skates, 65; dogfish, 321. 2. 15 miles off Monhegan, June 24-25, 1913. Total fish caught, 5,463; skates, 170. 3. 20 miles east of Cape Cod, November 11, 1913. Total fish caught, 6,532; skates, 202. 4. Jeffreys Ledge, December 11-12, 1913. Total fish caught, 3,996; skates, 62. Now and then a long line comes in with a skate on almost every hook, but this is unusual. Fishermen report them as present on the inshore as well as the offshore fishing grounds throughout the year. On our seaboard skates are salable only in special markets and are of so Uttle commercial importance that in 1919, which may serve as a representative year, the total amount brought into the several ports of Maine, New Hampshire, and Massachusetts was only 102,739 pounds, valued at $550. From time to time a few have been utilized as fertilizer. All others caught are thrown overboard. All our common Gulf of Maine skates are oviparous, laying large eggs with blackish or sea-green leathery sheUs, roughly oblong in outline with a hollow tendril at each corner by which they cling to seaweeds. The empty egg sheUs — "mer- maids' purses" — are common on our beaches among the fiotsam along high-water mark. While still in the egg the embryos develop temporary external gill filaments from the walls of the giU clefts, but these disappear completely after hatching. Many years ago Wyman (1867) published some notes on the development of one of our local skates (species not named) and figured the newly hatched young, since which time no attention has been paid to the development or life history of any of the species that occur in the Gulf of Maine. Probably, however, all spawn over a considerable part of the year with an incubation period of from 4 to 8 months, as is true of most of the European skates."" The sting rays are viviparous, but it is not likely that any of these strays from the south breed in the Gulf of Maine. It is easy to tell a skate from a ray (at least among species with which we are concerned) by the presence or absence of a dorsal spine on the tail, while its large caudal fin places the torpedo at a glance, but identification of the several skates is proverbially difficult. In the following key we have endeavored to facilitate it by characters obvious in handhng them at sea or on the dock. KEY TO GULF OF MAINE SKATES AND RAYS 1. No lopg dorsal spine on the tail . 2 Tail with long dorsal spines (sting rays) 8 2. Two small dorsal fins, but no distinct caudal on the tail (includes all our common skates) . 3 There is a large triangular caudal fin as well as the two dorsals on the tail.. Torpedo, p. 68 3. The midline of the ba«k, immediately over the backbone behind the shoulders, does not bear a row of large thorns, though it may be flanked by such 4 The midline of the back bears a row of large thorns on the rear part of the disk, on the tail, or on both 5 <• Clark (Journal of the Marine Biological Association of the United Kingdom, New Series, Vol. XII, No. 4, Oct., 1922, p. 629) described the eggs and young fry of several British species. 58 BULLETIN OF THE BUREAU OF FISHERIES 4. Teeth in about 50 rows; upper surface brown with dark spots Little skate, p. 58 Teeth in about 90 rows; upper surface usually with two large whitish eye spots near the rear angles of the disk'" Spotted skate, p. 60 5. The front angle of the disk is much blunter than a right angle; the whole upper surface of the disk is more or less thorny, with a row of very large thorns along the midline behind the shoulders Prickly skate, p. 62 Front angle of the disk is not blunter than a right angle; smoother species with notice- able thorns only in restricted patches 6 6. Front angle roughly a right angle with the snout hardly projecting; with stout thorns on the midline of the disk as well as of the taU 7 Front angle more acute than a right angle, with the blunt tipped snout projecting; no thorns in the midline except on the tail Barn-door skate, p. 66 7. Tip of snout blunt; outer corners of disk bluntly angular; thorns large Brier skate, p. 64 Tip of snout sharp-pointed; outer corners of disk rounded; thorns small Smooth skate. Raja senta, p. 65 8. No dorsal fins on tail 9 Tail with a dorsal fin in front of spine Cow-nosed ray, p. 72 9. Tail rounded above, without a keel Sting ray {Dasybatus marinus), p. 70 Upper side of tail, behind the spine, with a distinct keel Sting ray (D. haslatus), p. 70 THE SKATES. FAMILY RAJID^ 20. Little skate {Raja erinacea Mitchill) Common skate; Bonnet skate; Summer skate; Hedgehog skate; Old mahj; Tobacco box Jordan and Evermann, 1896-1900, p. 68. Garman, 1913, p. 337. Description. — The most diagnostic characters are small size, absence of thorns along the midline of the back, and blunt nose. The anterior angle of the disk is blunter than a right angle and the tip of the snout is rounded, with margins biilging opposite the eyes. The teeth are in about 50 rows. Females have thorns scattered all over the upper surface except on the midline back of the shoulder girdle, espe- cially prominent on head, snout, shoulders, and sides of tail. Males are less spiny, but in both sexes the spines on tail, shoulders, and along either side of the back ridge are especially strong. Males have bands of erectile hooks near the outer cor- ners of the pectoral fins, presumably for holding the female. The outer angles of the pectorals are bluntly angular. The two dorsal fins are close together; the tail is about half the total length. Size. — 16 to 20 inches; at the most 2 feet in length. Northern specimens average larger than southern. A specimen 20 inches long is about 12 inches wide. Color. — ^Grayish to dark brown, or clouded light and dark brown above, paler at the edges of pectoral fins; usually with many small round darker spots; white or grayish below. General range. — Coastal waters off the Atlantic coast of America; Nova Scotia and Gulf of St. Lawrence to Virginia. Occurrence in tJie Gulf of Maine. — This, the smallest of our skates, is the com- monest and the most familiar from its habit of coming up into very shoal water in ^ When this eye spot is laclving, as sometimes happens, it may be necessary to count the teeth to separate the " spotted ' ' from the "little " skate. FISHES OF THE GULF OF MAINE 59 sTommer and of stranding on the beaches, where dried skate carcasses are often to be' seen. It occurs all along the coasts from the Gulf of St. Lawrence and Nova Scotia to Cape Cod, and much farther south. It is very abundant both on the New Brunswick and the Scotian sides of the Bay of Fundy, and is taken everywhere and anywhere along the coasts of Maine and Massachusetts, far more commonly, in- deed, than one might suspect from the few definite records that have found their way into scientific literature. To what extent it enters into the skate population ofjthe offshore banks is as yet unknown. The little skate carries out a more or less definite migration up into shoal water in April and May, where it remains throughout the summer, autimin, and early winter, to return again to somewhat deeper water, say 30 to 50 fathoms, in Decem- FiG. 22.— Little skate (.Raja erinacea). After Garman ber or January. In summer it is perhaps most numerous at depths of from 5 to 15 fathoms, many even following the shelving bottom up to within a few feet of low-water mark. Others, however, lie deeper. It has been trawled at 25 fathoms even in midsummer, for example. On Georges Bank it is probably to be found at 30 to 40 fathoms throughout the year, and there is no reason to suppose that it ever descends to any greater depth than this. It is common knowledge that skates are most abundant on sandy or pebbly bottom; however, they are likewise found on mud and over ledges. They bite the hook readily, affording amusement to vacationists. Food. — Little skates are omnivorous. Hermit and other crabs, shrimps, worms, amphipods, ascidians ("sea squirts"), bivalve moUusks, squid, small fishes, and even such tiny objects as copepods have been found in their stomachs. Prob- 60 BULLETIN OF THE BUREAU OF FISHERIES ably crabs loom largest in their diet, for more than 29 per cent of the skates opened by Field (1907, p. 26) contained them; 15 per cent had bottom-dwelling shrimps (Crago); 6 per cent had eaten squid. Launce, alewives, herring, cunners, silver- sides, tomcod, silver hake, all have been found in the stomachs of these skates. Habits. — The spawning habits of the summer skate have not been followed in the Gulf of Maine, but oS southern New England its eggs have been taken as early as March and in abundance during July, August, and September, both [in fish traps and in dredges in a few fathoms of water. In all probabihty its breeding covers the same period north of Cape Cod — that is, eggs are laid in spring and earlj- summer, hatching in late summer and autumn. The eggs measure about 2 by 2J^ Fig. 23.— Big skate (Raja diaphanis). After Garman inches, and the great majority of the empty skate eggs washed up on the beach belong to this species. Huntsman's observations suggest that young hatched near the head of the Bay of Fundy descend to deeper water the first winter, and this probably applies to the Gulf of Maine as a whole. 21. Big skate {Raja diapJinnes 'MitchiH) Spotted skate; Winter skate; Eted skate - Jordan and Evermann (Raja ocellata Mitchill), 1S96-1900, p. 68. Garman, 1913, p. 339. Discriftion. — This skate is much like the little skate, but is larger, has more numerous teeth, and is of a different color. The front angle of the disk is much FISHES OF THE GULF OF MAINE 61 blunter than a right angle, bulging opposite the eyes, and the tip of the snout is rounded. The teeth are in from 80 to 110 rows on a jaw instead of in only about 50 rows, as in erinacea, and they are sharper in males than in females. The backs of both sexes are rough, with sharp spines on the head, around the eyes, along the anterior margins of the pectorals, over the shoulders, and on the sides of the tail, but the midline of the back behind the shoulders is free of spines, at least in adults. Males have rows of retractile hooks on the outer parts of the pectorals. The two dorsal fins are close together; the outer corners of the pectorals are bluntly angular; the claspers in males reach about halfway back along the tail, which occupies about half the total length of the fish. Size. — This skate grows to about 3 feet in length, commonly from 30 to 34 inches; specimens 32 inches in length are about 21 inches wide. Color. — Light brown above with round darker brown spots. As a rule there is a large white eye spot with black center near the posterior angle of the pectoral fin, and often two smaller ones close to the latter. When these eye spots are present they serve to identify this skate at a glance; sometimes, however, they are lacking, in which case half-grown specimens so closely resemble the little skate that recourse must be had to the number of teeth to tell one from the other. There is a translucent or white area on each side of the snout in front of the eyes and the lower surface is white. General range. — Atlantic coast of North America from New York northward to the Gulf of St. Lawrence, where it is common. Occurrence in the Gulf of Maine. — This, the second in size of our skates, occurs commonly all around the Gulf of Maine from Nova Scotia to Cape Cod. There are many locality records from the Bay of Fundy as well as from the coasts of Maine and Massachusetts, and it probably makes up a large proportion of the skate population on Georges Bank. It is very plentiful in Massachusetts Baj', but so closely does a two-thirds grown big skate resemble the adult little skate (p. 58) that it is often impossible to tell to which species reports refer. It is said to come up into shoal water on sandy beaches, but we have no first-hand information to offer on this point, and at Woods Hole it is never found in water shoaler than 5 to 6 fathoms. South of Cape Cod the name "winter skate" is appropriate enough, for it is only during the cold season that it is common about Woods Hole. Similarly, it is said to be taken in larger niunbers in winter than in suimmer in the Massachu- setts Bay region, though we can not verify this. However, this is distinctly a mis- nomer in the northern part of the Gulf of Maine, for not only is it abundant in shoal water in the Bay of Fundy (e. g., Passamaquoddy Bay) from May to Novem- ber, but to judge from temperature this probably applies to the whole coast line east of Cape Elizabeth. Spotted skates feed on the same diet as do little skates. Rock crabs and squid are their chief diet, but they also take annelids, amphipods, shrimps, and razor clams, and they prey upon whatever small fish are available, the list at Woods Hole including smaller skates, eels, herring, alewives, bluebacks, menhaden, smelt, launce, chub mackerel, butterfish, cunners, sculpins, silver hake, tomcod, and hake." 61 From Vinal Edwards' and Linton's notes. 62 BULLETIN or THE BUREAU OF EISHEEIES This skate is taken on hook and line, in weirs, and in otter trawls. The breed- ing habits of this species, as they apply to the Gulf of Maine, have not been traced. Its egg cases are little larger than those of the little skates — 2J^ by 1% inches. 22. Prickly skate {Raja scabrata Garman) Jordan and Evermann (Raja radiala Donovan), 1896-1900, p. 69. Garman, 1913, p. 340. Description. — The prickly skate can be identified at a glance, or rather touch, by the fact that the midline of the back behind the shoulders, and of the tail, is armed with a row of very stout thorns. As in the little and spotted skates, the Fig. 24. — Prickly skate {Raja scabrata). After Garman anterior angle of the disk is blunter than a right angle, its margin bulging some- what abreast of the eyes, and the tip of the snout is blunt. There is a pair of large, hooked tubercles or bucklers on each shoulder, one in front of and one behind each eye, as well as one behind each spiracle, besides the mid-dorsal row of 14 or more just mentioned. Smaller thorns occur on the snout and are scattered generally over the upper surface of the pectoral fins. The bases of the spines on the pectorals are star-shaped, a very diagnostic character; those of the bucklers shieldlike. Males ■ have two rows of hooked, erectile thorns near the outer corners of the pectorals, the latter being more angular than in either the little or spotted skates, while the two FISHES OF THE GULF OF MAINE 63 dorsal fins are separated by a definite space in the adult but are confluent in the young. Size. — This skate grows to 23^ feet in length, or slightly larger; males as small as 26 inches, nearly mature, have been found. Specimens 21 to 22 inches long are 15 to 16 inches wide. Color. — Brown above, either uniform or slightly clouded with Ughter and darker. Young ones are spotted with darker brown, but adults ordinarily lack these spots. Garman (1913, p. 34) mentions a partial albino, white above with a few reddish-brown and brown spots. General range. — The prickly skate is a northern cold-water fish, its range hardly extending west or south of Cape Cod, for it appears but rarely and at long intervals at Woods Hole, nor is it known south of tliis. How far north it ranges is yet to be determined. It is plentiful along the east coast of Nova Scotia and in the Gulf of St. Lawrence where it lives indifferently on the ice-cold banks and in the warmer water in the bottom of the deep channels, but it has not been recorded from Labrador north of the Straits of Belle Isle nor so far as we can learn from the eastern shores of Newfoundland. In north European waters it is represented by an extremely closely allied if not identical form {Raja radiata), which occurs from the Bay of Biscay in the south to Greenland, Spitzbergen, and the White Sea in the north. Occurrence in the Gulf of Maine. — The prickly skate is usually thought to be less common on our coast than either of the two species just mentioned, and it is certainly rare in very shallow water within our limits; but it is frequently taken on the New Brunswick side of the Bay of Fundy in depths of 10 fathoms or deeper, in 20 to 30 fathoms in St. Mary Bay (Nova Scotia)," while we ourselves trawled it (13 specimens) in 22 and 27 fathoms on sandy bottom in Ipswich Bay in July, 1913. Since it has also been recorded from Casco Bay, Ipswich Bay, Gloucester, Salem, Nahant, and Provincetown, it evidently occurs generally all along the shores of the Gulf in moderate depths. Judging from the considerable depths to which its European relative descends — it has been trawled down to 450 fathoms — skates caught in the deeper parts of the Gulf are more likely to belong to this than to either of the preceding species, and it may be the prevalent skate on the oflfshore banks. It has not been recorded below about 200 fathoms off our coasts. Habits. — Notliing is recorded of its habits in the Gulf, nor, so far as we can learn, have its eggs or young ever been definitely recognized there, but probably what is known of the spawning habits of its European representative appUes equally here, briefly, that it comes up from deeper water into shoal water in spring to spawn there during the summer, retreating once more to greater depths in winter; that the egg case measures about 2V2 by 1 Y5 inches (exclusive of its tendrils) ; and that the fry remain near land during their first winter. Food. — The pricldy skate, like most of its relatives, feeds indiscriminately on small fish, ampliipods, worms, etc. Such, at least, is true of the European form. So far as we can learn no stomachs have been examined on this side of the Atlantic. " According to Huntsman (1922a'). 102274— 25t 5 64 BULLETIN OF THE BUREAU OF FISHERIES 23. Brier skate {Raja eglanteria Boso) Jordan and Evermann, 1896-1900, p. 71. Garman, 1913, p. 341. Description. — The brier, like the piickl}^ skate, is armed with a row of stout thorns along the midline of the back from shoulder to dorsal fin near the tip of tail. Otherwise, however, it is a much smoother species and its snout is more acute. There are groups of large spines opposite and behind the eyes and on the sides of the tail, with a pair on each shoulder. Elsewhere the upper surface of the disk Fio. 25.— Brier skate (,Raja eglanteria). Alter Garman bears only very small but very sharp prickles, these being most numerous on the anterior parts of the pectorals, over the head and snout, and on the middle of the back and tail among the larger thorns, whence its common name. The males, we might add, are provided with several rows of large erectile hooks on the outer parts of the pectorals, which the females lack. The snout angle is roughly a right angle, its margin bulging less opposite the eyes than in any of the blunter-nosed FISHES OF THE GULF OF MAINE 65 skates. The outer corners of the pectorals are distinctly angular. The dorsal fins are separated by a considerable interval in which there are usually one or two spines, instead of close together as in the httle, spotted, and prickly skates. Size. — The brier skate grows to a length of about 2 feet. Specimens of from 21 to 22 inches are 13 to 14 inches wide. Color. — Described as bro'wn above; the pectorals variously mottled, blotched, and barred with darker; a translucent or white space on each side of the snout; white below. General range. — Off the eastern coast of the United States from Cape Cod to Florida. Occurrence in the Gulf of Maine.— This is a southern species, uncommon even as far north as Woods Hole and decidedly rare in the Gulf of Maine, where it has been taken only at Provincetown and at Gloucester, the latter its most northerly outpost. Fig. 26.— Smooth skate (itaja scnta). After Garman 24. Smooth skate {Raja senta Garman) Jordan and Evermann, 1896-1900, p. 71. Garman, 1913, p. 338. Description. — This skate is recognizable by its sharp snout, the rounded out- line of the outer margins of the pectoral fins, and by the fact that the mid-dorsal line of thorns runs back only to about the middle of the tail, where it dwindles and disappears. There are also large spines on the front parts of the pectoral fins, on the ridges about the eyes, and a group on each shoulder; otherwise the back and top 66 * BULLETIN OF THE BUREAU OF FISHEEIES of tail are rough with small spines only, except that males have the usual rows of hooks on the outer parts of the pectorals. The anterior snout angle is roughly a right angle; the tip of the snout itself is sharp instead of rounded. The two dorsals are close together, not separated by spines as in the brier skate. Size. — The largest recorded specimen was 22 3^^ inches long, the tail being almost exactly half the total length. Its width was 14 inches. Color. — Rusty brown above clouded with darker, not spotted; no doubt white below like other skates. Young examples have been seen with white spots. General range and occurrence in the Gulf of Maine. — Of the distribution of this skate, evidently very rare in our Gulf, nothing is known further than that speci- mens have been taken on LaHave Bank and off Provincetown, on the strength of which it has usually been described as "a deep-water form. Banks of Newfoimdland Cape Cod." Nothing whatever is known of its habits. :%■ /:-■ / -:' M ^^^ v t- Fig. 27.— Barn-door skate {Raja stabuliforis) 25. Barn-door skate {Raja stabuliforis Garman) Jordan and Evermann, 1896-1900 {Raja Imvis Mitchill), p. 71. Garman, 1913, p. 341. Description. — The barn-door skate is easily identified by its large size, very pointed snout, and smooth skin. The mid-dorsal thorns are comparatively small and 'run only from the hinder part of the back over the tail; otherwise the spines, which are very small, are restricted to the sides of the tail, top of the tip of the snout, and to narrow bands along the front edges of the pectoral fins, in front of and between the eyes, with a few scattered here and there over shoulders and back. Thus the whole upper surface is smoother than in any of the other skates. The male is provided with the erectile hooks on the outer parts of the pectorals common FISHES OF THE GULF OF MAINE 67 to all skates of this genus. The anterior angle of the disk is sharper than in other skates, being more acute than a right angle. The snout itself is long but blunt tipped, the outer corners of the pectorals are angular, and the disk as a whole is diamond or lozenge shaped. The two dorsal fins are separated by a short space, with one or more spines, and the tip of the tail extends farther beyond the second dorsal than in most skates. Size. — The barn-door is our largest skate, growing to a length of 6 feet or even more. One of 58 inches length was 42 inches wide with a tail 27 inches long.^' Color. — The barn-door, like so many sea fish, varies in color. As a rule the upper surface is bro^vn, usually of a distinctly reddish hue, variously marked with small scattered darker spots or blotches of varying size, often with pale marblings or waterings. The lower surface is not as uniformly pale as in most skates, its gray or white ground being shaded with darker toward the snout and speckled with black over the abdomen. General range. — ^Atlantic coast of North America from the Gulf of St. Lawrence and the outer coast of Nova Scotia, where it is common, to Florida. In European seas it is replaced by a very close ally, the "common skate," Baja batis. Occurrence in the Gulf of Maine. — This is a common fish off the New England coast and in all parts of the Gulf. Any very large skate taken or reported there is almost certain to be a "barn-door." Following the coast around from east to west we find it reported as plentiful off the Nova Scotian shore; it is knowTi from St. Mary Bay; is found very generally though not abundantly in the Bay of Fundy and up in Passamaquoddy Bay; is reported from Eastport, Casco Bay, and generally along the coast of Maine; is known from various localities in Massachusetts Bay, where we have seen many caught; and is taken in abundance by the trawlers on Georges Bank.^^ In short, it is to be expected anywhere in the Gulf. Like most other skates, it is often taken in shoal water in summer; seldom or never in winter. Huntsman tells us that it comes up into Passamaquoddy Bay from May to Novem- ber. We took one nearly 5 feet long at Cohasset in Massachusetts Bay in only a couple of fathoms of water in midsummer. Indeed, it is often stranded on the beaches. This inshore migration, however, does not involve the entire stock — witness its presence in 20 to 60 fathoms on Georges Bank and off Cape Cod throughout the year and the fact that it is reported by fishermen and has been trawled by vessels of the bureau below 100 fathoms in smnmer. In the warmer waters off the south coast of New England it comes inshore in spring and autiman, descending to deeper water in summer. Habits and food. — Barn-door skates, like other skates, are bottom swimmers, preferring smooth to rocky ground, but the fact that the lower surface is more or less pigmented instead of white suggests that it hugs the bottom less closely than do other skates. Garman, the foremost authority on this group, has pointed out that the spines on the snout of this skate are usually worn smooth, as though used to dig in the mud or sand — very likely it thus obtains the bivalves that form part of its diet. It also feeds on worms, various crustaceans, particularly large rock crabs " Described by Garman (1913, p. 342). 68 BULLETIN OF THE BUREAU OF FISHERIES and lobsters, on squid, and on fish. Probably, thanks to its large size, it is more destructive to the latter than are any other skates. At Woods Hole the list includes spiny dogfish, alewives, herring, butterfish, launce, cunners, tautog, menhaden, sculpins, silver hake, hake, and flatfish. No doubt cod, haddock, etc., suffer to some extent from this skate on the offshore fishing grounds, for its European relative is a well-known enemy of the cod, and there is no reason to suppose that our "barn- door" is less voracious. It is a strong, active swimmer, as anyone who has landed a large one on a hand line will agree. It bites readily on almost any bait, and is often caught on hand and long lines as well as in the otter trawl and in weirs along shore. Breeding habits. — Very little is known of its breeding habits. Probably it spawns when in shoal water, that is, during the warm season of the j^ear. Eggs, probably belonging to the barn-door, are 5}4 by 2^ inches.^^ Commercial lvalue. — The barn-door skate is of no commercial value except as entering into the small landings of skates mentioned on page 57. \ ^r^'" Fig. 28. — Torpedo (Narcacion nobilianus) THE TORPEDOES. FAMILY NARCACIONTIDiE 26. Torpedo {Narcacion nohilianus Bonaparte) Electric skate; Crampfish; Numbfish Jordan and Evermann (Tetranarce occidentalis Storer), 1896-1900, p. 77. Garman, 1913, p. 310. Description. — No one would be apt to mistake the torpedo for anj^ other skate or ray, the rounded outline of the disk and the large caudal fin identifying it at a glance. Furthermore, the skin is soft and naked, without the spines so character- istic of all our common skates. The disk is roughl}'^ subcircular, truncate in front, and considerably broader than long. The eyes are very small and are set far for- K Doctor Qarman supplied this note. FISHES OF THE GULF OF MAINE 69 ward. The two dorsal fins, of which the anterior is the taller, stand at the forward end of the tail, the anterior, indeed, partly above the bases of the ventral fins, and they are separated by an interspace nearly as long as the second dorsal fin. The tail fin is of ordinary fish form — triangular and about three-quarters as long as deep. The tail is shorter than in the skates — that is, it occupies only about one- third the total length of the fish. The most interesting feature of the torpedo is its electric organ and its ability to give electric shocks of considerable strength to anyone touching it. Color. — Dark chocolate brown above; lower surface white except that the edges of disk, fins, and tail are of the same dark chocolate tint. Size. — Adult torpedoes are usually 2 to 5 feet long and heavy for their size. Specimens taken at Woods Hole average about 30 pounds, ranging from 4 or 5 up to 75 pounds. Torpedoes as heavy as 200 pounds have been recorded, and they have been taken up to 170 pounds or more in Massachusetts Bay. General range. — Tropical and temperate parts of both sides of the Atlantic; Maine to Cuba on the American coast. Occurrence in the Gulf of Maine. — -The torpedo is a southern fish, which, like so many others, finds the northerly limit to its comnnon occurrence at Cape Cod. It strays past the cape into the Gulf of Maine often enough, however, to be looked upon as a regular, if rare, summer visitor. It has been recorded even as far east as Eastport (not, however, in the Bay of Fimdy), as well as at various other localities along the coast of Maine (e. g., Williamsport, off Seguin Island, and Casco Bay); likewise at Cape Ann, in Cape Cod Bay, near Provincetown, and along the outer shores of Cape Cod, so it would be no surprise to find it anyAvhcre along our shores. It has also been reported from Georges Bank. Most of the records date back many years. It is said to have been unusually common in the year 1819 and for four or five years thereafter. Again in 1845 about a dozen came ashore or were otherwise taken near Provincetown. It is as apt to be found in our Gulf now as then, for one was seined off Seguin in aliout 1880, another was caught on a long line set from the Grampus on LaHave Bank in the siunmer of 1890, one was taken in a trap at Wood Island near Cape Elizabeth in 1894, and torpedoes were collected by Dr. W. C. Kendall of the Bureau of Fisheries at several localities along the coast of Maine in 1896. West of Cape Cod it is much mere numerous, appearing not uncommonly trom May to November about Woods Hole. Breeding and halits. — The torpedo, like others of its tril)e, is a bottom fish. It feeds chiefly on small fish and to some extent on Crustacea. Probably it does not succeed in breeding in the cold waters of the Gulf, but at Woods Hole it has been found to contain nearly ripe eggs by the end of June. It is viviparous, the embryos having been figured by Garman (1913, pi. 01). Commercial value. — Nowadays the torpedo is of no commercial value, but years ago before the use of kerosene oil was general its liver oil was considered equal to the best sperm for illuminating purposes. 70 BULLETIN OF THE BUEEATJ OF PISHEBIES THE STING RAYS. FAMILIES DASYBATID^ AND MYLIOBATID^ 27. Sting ray (Dasyhatus marinus Klein) Stingaree; Clam cracker Jordan and Evernaann (Dasyatis centrura Mitchill), 1896-1900, p. 83. Garman, 1913, p. 382. Description. — The most characteristic features of the sting ray are the very long whiplike tail without dorsal fins and the strong saw-toothed spines that the tail bears on its dorsal surface. The disk is roughly quadrangular, one-fourth wider than long, with the anterior corner much blunter than a right angle, the anterior and posterior margins nearly straight, and the lateral corners bluntly angular. The ventral fins are relatively much shorter than in the common skates. The tail is more than twice as long as the disk, rounded above and tapering regu- larly to a very narrow tip. The spines, of which there are from one to several,^' are situated about one-fifth of the way back along the tail. Young sting rays are smooth-skinned, but adults bear scattered tubercles on the middle and hind parts of the back and on the back and sides of the tail, which become more and more numerous as the ray grows. Size. — Maximum length, including the tail, about 12 feet. Color. — The general ground tint varies according to the background. General range. — Both sides of the tropical and temperate Atlantic, north on the American coast to Cape Cod, and (according to Smith ") not known south of Cape Hatteras. Occurrence in the Gulf of Maine. — The only claim of the sting ray to mention here is that it has been recorded from Chatham, on the outer shore of Cape Cod, and that it is said to have been seen on the shoaler parts of Georges Bank. It has no real status as a Gulf of Maine fish, where it appears only as a waif from the south, though common enough as far east as Woods Hole, where it appears in June or early July. Beware of handling any skate with a long whiplike tail lest it prove a sting ray, for its spine is a dangerous weapon. Breeding habits. — The sting rays are viviparous. 28. Sting ray {Basylatus Tiastatus DeKay) Jordan and Evermann {Dasyatis hastata DeKay), 1896-1900, p. 83. Garman, 1913, p. 391. Description. — This ray so closely resembles D. marinus that we need only point out that the tail bears a low keel on its dorsal surface behind the spines instead of being rounded above as in its relative. General range. — Atlantic coast of America, Cape Cod to Brazil. Occurrence in the Gulf of Maine. — This southern sting ray is mentioned here because recorded many years ago from Chatham on Cape Cod. M There are three in a specimen figured by Garman and two in one we have examined. " The fishes of North Carolina, by Hugh M. Smith. North Carolina Geological and Economic Survey, Vol. II, 1907, p. 44. FISHES OF THE GULF OF MAINE 71 FiQ. 29.— sting ray (Dasybatus marinus). After Qarman 102274— 25t 6 72 BULLETIN OF THE BUREAU OF FISHERIES 29. Cow-nosed ray {Rhinoptera quadriloba LeSiieur) Jordan and Evermann (Rhinoptera bonasus Mitchill), 1896-1900, p. 90. Garman, 1913, p. 444. Description. — The cow-nosed ray and all its close relatives are of such charac- teristic batlike outline, with head so peculiar and teeth so different from those of our other Gulf of Maine skates and rays, thatonceseen they are never apt to be mistaken for anything else. The anterior angle of the disk is much blunter than a right angle; the outer corners of the pectorals are acute, pointed, and their posterior margins distinctly concave. The ventral fins are comparatively very small, longer than wide, reaching but a short distance back of the posterior corner of the pectorals. There is a single small dorsal fin originating a short distance back of the bases of the ventrals, and Lnamediately back of it stands a stout spine. The tail is hardly twice as long as the disk, whiplike and tapering to a very slender tip. The cranimn Fig. 30. — Cow -nosed ray (Rhinoptera quadriloba). After Garnian of the cow-nosed ray is raised above the general level of the disk with the large eyes set lateral instead of dorsal, and in front of the fins instead of far back as in other skates and rays. Its teeth are flat and arranged like the bricks or tiles in a pavement in a manner more easily figured than described. Size. — The cow-nosed ray grows to a length of about 7 feet. In one about 3314 inches in total length the disk was 25 inches long by 32 J^ inches broad.^' Color. — Brown above; white below, except toward the outer corners of the pectoral fins where it is brownish. General range. — ^Atlantic coast of the United States, Nantucket to Florida. »' Described liy Radclifie (Bulletin, U. S. Bureau of Fisheries, Vol. XXXIV, 1914 (1916), p. 279). FISHES OF THE QVLF OF MAINE 73 Occurrence in the Gulf of Maine. — The cow-nosed ray has even less claun to be called a Gulf of Maine fish than have the sting rays just mentioned, for while it is often taken in the traps at Woods Hole — 145 in one day on one occasion — and is recorded from Nantucket, it has never actually been seen east or north of Cape Cod. Chimseroids. Subclass Holocephali THE CHIMiERAS. FAMILY CHIM-ERID^ The chimseras find their nearest affinities in the sharks but are separated from the latter by many important anatomic characters, the most obvious of which are the facts that there is no spiracle, there is but one gill opening on either side, the tail is symmetrical, and the gills are fringelike and free at the tips like those of bony fishes. In general aspect the chimseras remotely suggest the grenadiers (p. 467), but are easily separable from them by the location of the ventral fins, which are set far ^back uiider or behind the tips of the pectorals; by the fact that the fin on the back is separated by a deep notch into dorsal and caudal portions; by the very small eye; and by the large size of the pectoral fins, to list only the most obvious differences. There is no danger of confusing them with any other Gulf of Maine fishes, so curious is their appearance. Fia. 31.— Chimsera (Chimscra affinis', 30. Chimaera. {Chimsera affinis Capello) Jordan and Evermann, 1896-1900, p. 95. Description. — The chimtera is deepest (one-seventh to one-eighth as deep as long) just behind the gills, tapering gradually backward to a weak slender tail, and is very soft bodied. The head is short, its dorsal profile oblique and prolonged into a short, soft, conical knob above the mouth. The forehead of the male bears a curious cartilaginous hook, armed with recurved prickles on its lower surface, which probably serve to clasp the female. The mouth is inferior in position, relatively small, the upper jaw with four, the lower with two, flat plates, set edgewise, in place of teeth, and with thick fleshy hps. The gill openings are vertical, set low down on the sides of the neck, and each is covered with a flap of skin paralleling the gill covers of bony fishes. There are two distinct dorsal fins. The first of these originates over the gill opening, is triangular, about as high as long, and supported at its anterior margin by a stout spine that is free at the tip. The second dorsal is separated from the first by a space that probably varies in length, and is less than half as high as the 74 BULLETIN OF THE BUREAU OF FISHERIES first, with straight margin. The small caudal fin is demarked from the second dorsal by a deep notch; it is lanceolate in outline, terminates rearward as a short whipUke filament, and extends a short distance forward on the ventral surface of the trunk, there being no separate anal fin. The ventrals and pectorals are both triangular and pointed, the latter being much the larger and reaching back nearly to the point of origin of the ventrals. In the male the lower part of each ventral fin is modified as a trifid clasping organ. The skin is smooth, or perhaps shghtly prickly, the lateral Une well developed, ramifying in several branches over the head. This species^" is a close ally of the well-known chimasra of north Eiu'opean seas (C. monstrosa), but is distinguishable from it by the facts that it has no separate anal fin, that there is a considerable free space between its two dorsal fins, that the outline of the second dorsal fin is straight, that its caudal filament is much shorter, and that its pectorals hardly reach back to the ventrals. Color. — ^Leaden all over. Size. — Maximum length about 3 feet. General range. — Not uncommon on the continental slope of North America from the latitude of Cape Cod northward, in 300 to more than 900 fathoms. Occurrence in the Gulf of Maine. — We mention the chimjera here because one (or more) was brought in from Georges Bank some time between 1877 and 1880.°° It would be no surprise to find them on the seaward slope of the bank, for halibut fishermen have often caught them off LaHave and the more easterly banks. One has even been found in the harbor of Noank (Conn.), but there is no record of it in the inner parts of the Gulf of Maine. Habits and food. — Nothing whatever is known of the habits of this chimaera; little more of the northern Eui'opean species except that it is a groimd fish, omniv- orous, eating small fish, mollusks, Crustacea, echinoderms, and worms, and that it produces large eggs with horny oval cases, bearing threadlike filaments. The bony fishes. Subclass Teleostomi THE STURGEONS. FAMILY ACIPENSERID.S The sturgeons — the only Gulf of Maine representatives of the ganoid fishes — share with the sharks an uneven tail with the vertebral column extending out into the upper lobe, but there is no danger of taking one for a shark as there is but one gill opening on each side, and the gills are inclosed by bony gill covers. 31. Sturgeon {Acipenser sturio Linnaeus) Jordan and Evermann, 1896-1900, p. 105. Description. — Sturgeons are easily distinguished from all our other salt-water fishes by the fact that the head is covered by bony plates united by sutures, and the skin is armored by a row of large bony shields or bucklers along the mid-back, - *• This flsh is generally considered identical with a chimaera taken ofl the coast of Portugal, hence the choice of the specific ame affims instead of plumbea, by which the chimaera of North American waters was first known. •* Report, U. S. Commission of Fisheries, 1879 (1872). p. 788. FISHES OF THE GULF OF MAINE 75 with two rows on each side, while the tail is of the "shark type," that is, with the iixis bent upward and upper lobe much longer than the lower, and the nose is long and curiously depressed. Each buckler bears a longitudinal keel and spine, those of the dorsal series being much larger than the others. On the average there are about 11 (10 to 16) bucklers in the dorsal row, 28 (26 to 34) in each upper lateral row, and 9 to 14 in each lower lateral row. The dorsal row extends from over the gill cover to the dorsal fin; the upper lateral from the corner of the gill opening back to the base of the tail fin; the lower lateral row from immediately behind the pectoral fin to the ventral fin, and again from the latter to the anal fin. The single rather small triangular dorsal fin is far back, its hind edge over that of the still smaller anal. The ventrals are likewise far back. The pectorals are set almost as low as the plane of the belly. The body is elongate, comparatively slender and more or less pentagonal in cross section owing to the rows of bucklers, instead of rounded as in most bony fishes. In large fish the snout is about one-third the total length of the head (longer, comparatively, in small ones), depressed below the level of the forehead, and nearly flat beneath. The mouth, which is situated on the under side of the snout, is small and toothless (except in larval stages), with protractile lobed lips, and there are four pointed barbels in a row across the lower surface of the snout in front of the mouth. Fig. 32. — Sturgeon (AcipeiiseT stand) Color. — Olive greenish or bluish gray above (in some seas reddish above), gradually fading on the sides and changing rather abruptly below the upper lateral row of shields to the white of the belly. Size. — The sturgeon is a very lai'ge fish, specimens as long as 18 feet having been recorded from Europe and from New England; nor are 10-foot cows uncom- mon to-day in the Delaware River, where sturgeons are more plentiful than they are anywhere in New England. As a rule adults taken there run from 6 to 10 feet in length, with about 7 feet as the maximum for the males. Females weigh up to 350 pounds or more, while males average about 65 pounds in weight. Sturgeons (male and females together) averaged about 120 pounds in the Kennebec during the years when the fishery was carried on there. The fact that a fish between 11 and 12 feet long, taken near Helgoland in the North Sea, weighed 623 pounds will give an idea of the weight they sometimes attain. General range .^Yioth. sides of the North Atlantic, from Scandinavia to the Mediterranean on the European coast and from the St. Lawrence River to the Gulf of Mexico on the American coast." " Also recorded by Prince from Hudson Bay (Report ol the sixty-seventh meeting of the British Association lor the Advance- ment of Science, held at Toronto in August, 1897, p. 687) . 76 BULLETIN OF THE BUREAU OF FISHERIES Occurrence in the Gulf of Maine. — The sturgeon, like the salmon, shad, and ale- wife, is anadromous, spending most of its life and making most of its growth in the sea or in bays and estuaries, but running up the larger rivers to spawn. During its sojourn in salt water it may be expected anjrwhere oflE the coasts of the Gulf of Maine. There is definite record of it at sundry localities on both sides of the Bay of Fundy; in Penobscot Bay; Casco Bay; at the mouth of the Piscataqua River; on the Boar's Head — Isles of Shoals fishing ground, where several of from 3 to 4 feet were taken in the gill nets during April and May, 1913; both outside and in Boston harbor; off Provincetown; and off Truro on Cape Cod. It is well known in the St. John, Penobscot, Kennebec, and Merrimac Rivers, and has even been taken some distance up as small a stream as the Charles River. In fact, sturgeon once entered practically every stream of any size emptying into the GuK of Maine. Writing of Massachusetts in 1634, Wood (1634, p. 37) described the sturgeon as "all over the countrey, but best catching of them be upon the shoales of Cape Codde and in the river of Merrimacke, where much is taken, pickled, and brought for England, some of these be 12, 14, and 18 foote long." It is only the comparative rarity of the sturgeon in the GuK of Maine that limits its commercial importance in the tributary rivers. In the year 1919 a total of 20,227 pounds was landed in Maine and Massachusetts. Supposing each car- cass to have weighed 50 pounds (a very low estimate) , this means a total of 400 fish at the most. In former years, when our rivers were less obstructed and the stui'geons more plentiful, a fishery was intermittently maintained in the Kennebec, but as far back as the year 1880 the catch for the year was only 250 fish yielding 12,500 pounds of meat. So far as we can learn the open GuK has never supported a sturgeon fishery, the few taken there being picked up accidentaUy in drKt nets, traps, or weirs."' Habits. — Although much attention has been paid to the sturgeon in other parts of the world because of its great economic importance, knowledge of its movements in the Gulf of Maine is of the scantiest. The large adult fish enter the mouths of our rivers sometime late in the spring in company with the salmon, shad, and alewives, slowly working theu- way upstream beyond tidewater before they deposit their eggs. Spawning takes place, so far as known, in May, June, and perhaps as late as July. Such, at least, is its season in north European waters, and what little is on record of its movements in northern New England is of the same tenor. It has been suggested that some may spawn in brackish water, with which the fact that females with large eggs have been taken about Woods Hole in June and July (that is, at the height of the spawning season), is in accord. A single female fish may produce as many as 2,400,000 eggs, which hatch in about a week "^ after fertilization. Judging from European experiences with artificially reared sturgeon of this species, the larvae may be expected to reach a length of 12 mm. within five days after hatching; 16 to 17 mm. at two weeks; 20 mm. at four weeks; and 4 to 53^ inches at two months. 81 The short-nosed sturgeon (.AcipeiiscT bTevirostrum LeSueur) was reported from Boston harbor and from Hockport, Mass., many years ago, but it is probable that the specimens in question were small common sturgeons. " Ryder (Bulletin, United States Fish Commission, Vol. VIII, 1888 (1890), p. 231) has given an account of thespawning habi I : and early development of the sturgeon. FISHES OF THE GULF OF MAINE 77 The old fish return to the sea after spawning, but as yet we have no idea how many years in succession a given fish may spawn or to what age sturgeons live. Apparently some of the young fish take to the sea at one year while others live for two or three years in the rivers in which they are hatched ; for while sperlets of from 5 to 6 inches in length are found at the mouth of the Delaware River, as well as in the Elbe on the other side of the Atlantic,"** and in the North Sea, young sturgeon as long as 18 to 20 inches are taken in winter both in the Delaware and in the Merrimac Rivers. Three marked fish in the Elbe were found to have grown as follows: The first from 17 to 38 cm. from June 17 until the following April; the second from 43 H to 64 cm. from April 9 until the followdng December; and the third not at all from November until the following February, suggesting that the sturgeon, like many other fishes, makes most of its growth during the warm months. By the time they have reached a length of 3 feet or so all sturgeons are either in the sea or about the river mouths, and few of them become sexually mature until they grow to about 4 feet or more in length. How long a period is covered by this growth is not known, but immature fish of from 3 to 4 feet in length are common about Woods Hole throughout the summer season, at the time w"hen the lai-ger ones are in the rivers spawning. We have yet to learn how far oft'shore sturgeons stray. They certainly descend to at least 25 fathoms, for they have been caught on cod and had- dock lines at that depth in Scandinavian waters. The sturgeon is a bottom feeder, most abundant on sandy ground (such, at least, being the case in the North Sea) , swimming slowly to and fro when at peace but capable of darting ahead like an arrow on occasion, and frequently coming up to the surface to jump clear of the water. Though so sluggish that it usually offers no resistance when netted, large ones are very strong. An old North Sea proverb has it that leaping sturgeons and dancing girls are both hard to hold ! The adult sturgeon is a mud grubber, rooting in the sand or mud with its snout like a pig (the barbels serving as organs of touch) , as it noses up the worms and mol- lusks on which it feeds and which it sucks into its toothless mouth with considerable amounts of mud. It also consumes small fishes, particularly sand launce. Small ones, while living about estuaries and river mouths, subsist on amphipod and isopod Crustacea. Sturgeon, like salmon, eat little or nothing when running upriver to spawn. THE EELS. FAMILIES ANGUILLID.ffi, SYNAPHOBRANCHIDiE, LEPTOCEPHALIDiE, SIMENCHELYID.ffi, AND NEMICHTHYID.aE Eels have no ventral fins; scales are either absent or so small as to be hardly visi- ble; their fins are soft without spines; the gill openings are very small; the vertebrae extend in a straight line to the tip of the tail; and a single fin runs over the back and forward on the belly with no separation into dorsal, caudal, and ventral portions. There are several other fishes of eel-like form in the Gulf of Maine, viz, the hags and lampreys, rock eel (Pholis), snake blenny (Lumpenus), wryniouth (Crypta- canthodes), eel pout (Zoarces), and sand eel (Ammodytes), but the jawless, sucker- " Prince records a 6-inch sturgeon from Hudson Bay (Report or the sixty-seventh meeting or the British Association for the Advancement of Science, held at Toronto in August, 1897, p. 687) . 78 BULLETIN OF THE BUREAU OP FISHEKIES like mouth of the first two separates them at a glance from the true eels, wliile in all the rest either there is a well-marked separation between anal and caudal fins, ventral fins (large or small) are present, or the dorsal fin is spiny, not soft. Only four true eels are known from the Gulf of Maine: The commion eel (p. 78), the slime eel (p. 83), the conger (p. 8(3), and the snipe eel (p. 88), which fall into four different families according to American usage. Most European students, however, put the first three (with many other species) into one family (the Anguillidse), and the snipe eel and its relatives in a second (the Nemicthyidfe). A fifth species, the long-nosed eel (a deep-water form), is to be expected in the deepest parts of the GuK though it has not been recorded there as yet. The group likewise includes the trop- ical morays and sundry deep-sea forms, some of them exceedingly bizarre in appearance. Common, conger, and slime eels look much alike, but are separated from one another by the size of the mouth and by the relative lengths of the fins. In the snipe eels the two jaws are prolonged into a very long slender beak, recalling that of a silver gar, the tail is whiplike, the neck noticeably slimmer than the head, and the general form extremely slender. KEY TO GULF OF MAINE EELS 1. Both jaws are prolonged into a long slender bill Snipe eel, p. 88 The jaws are not bill-like, the snout being short and blunt 2 2. The anal fin originates well in front of the point of origin of the dorsal Long-nosed eel, p. 84 The anal fin originates behind the point of origin of the dorsal 3 3. The dorsal fin originates far behind the tip of the pectoral Common eel, p. 78 The dorsal fin originates close behind the tip of the pectoral 4 4. Mouth large, gaping back as far as the middle of the eye; very strong and active Conger, p. 86 Mouth small, falling short of the eye; soft, slimy, and feeble Slime eel, p. 83 32. Eel {Anguilla rostrata LeSueur) American eel; Common eel; Silver eel; Feesh-water eel Jordan and Evermann {A. chrysypa Rafinesque), 1896-1900, p. 348. Description. — In the common American eel the dorsal fin originates far behind the pectoral, this character being of itself a suflicient field mark to distinguish it from the conger, from which it also differs in that the lower jaw projects beyond the upper or at least equals it in length, and that its eyes are small and round. Further- more, it develops scales after it is about 3 or 4 years old, though these are so small that they might be overlooked. Like the conger, however, it has a pointed snout, a large mouth gaping back to or past the middle of the eye, and gill slits set horizontal on the sides of the neck, their upper corners abreast of the center of the base of the pectoral fin. It is very closely related- to the European eel {Anguilla vulgaris), but has fewer vertebra (about 107 as compared with about 114 or 115). Size. — Eels are said to grow to 4 feet in length and to 163/^ pounds in weight. Full-grown females average only about 2 to 3K feet, however, and males are smaller. Any eel more than 18 inches long would probably be a female, and one more than 24 inches in length would certainly be one. FISHES OF THE GULF OF MAINE 79 Color. — The color varies widely with the bottom on which the eel lives. As a rule it is dark muddy or olive-brown above, more or less tinged with yellow on the sides; the lower siu^ace paler brown and yellower, with dirty yellowish-white^belly. General range. — The area of distribution of the eel is peculiar. It lives most of its life and makes most of its growth in the estuaries and fresh rivers tributary a 7i&^*^ Fig. 33.— Common eel (AnguUla Tostrata) a, Adult. 6, Leptocephalus stage, 49 millimeters, c, Leptocephalus stage, 55 millimeters. After Schmidt, d, Leptoceph- alus stage, 58 millimeters. After Schmidt, e, Transformation stage, 61 millimeters. After Schmidt. to the east coast of America, from the Gulf of St. Lawrence south to the GuK of Mexico and to Brazil, and is common in the West Indies and at Bermuda. It moves out to sea to spawn, however, as described hereafter (p. 81). 80 BULLETIN OF THE BUREAU OF FISHERIES Occurrence in the Gulf of Maine. — The occurrence of the eel around the periphery of our Gulf can be described in the one word — "universal." From Cape Sable on the east to Cape Cod on the west there is, we believe, not a single stream mouth, muddy estuary, or tidal marsh into which fresh water empties but supports eels to some extent. They run up every stream, large or small, from which they even- tually find their way into the ponds at the headwaters unless barred by insurmount- able barriers such as very high falls. Eels are even caught in certain ponds without outlets. Habits. — Up to very recently the life history of the eel remained a mystery, for although the fact that the young "elvers" run up into fresh water in spring while adults journey downstream in autumn has been common knowledge for centm'ies, and while a host of myths grew up to explain the utter absence of ripe eels of either sex, either in fresh water or along the seacoast, it was not until a few years ago that the riddle of its breeding place was solved. Now we know that both the American and the European eel lives a life that is just the antithesis of that of the salmon, shad, or alewife, and, thanks chiefly to the persevering re- searches of the Danish scientist, Johannes Schmidt, the spa^vning grounds of both eels have been discovered and the history of their larvse traced."^ Briefly, the life history of the eel is as follows: The young "elvers" appear along our shores in spring when they average from 2 to 3J^ inches in length. As yet we have little data on the exact date of arrival, but they have been taken as early as March at Woods Hole and in April in Passamaquoddy Bay. The fact that elvers run up the streams emptying into Narragansett Bay from mid-April until mid-May, and that Welsh observed a tremendous run in Little River, near Gloucester, on May 5, 1913, suggests that they may be expected in the mouths of most Gulf of Maine streams late in spring. In the Bay of Fundy region, however, probably owing to the dilatory warming of the local streams, they are found ascending streams during the summer. The run in one stream may last for a month or more, while it may last only for a few days in another. Even at this early stage there is a noticeable habit of segregation, some remaining in tidal marshes and other estuarine situations while many go into fresh water, and some of them ascend the larger rivers for tremendous distances. It is now generally believed that only the females run up above the head of tide — that is, that any eel caught in fresh water is a female — but the evidence on which this assumption rests is none too conclusive, especially in the case of the American It is no wonder that the ability of the elvers to surmount obstacles on their journey upstream is proverbial, for they clamber over falls, up dams, etc., and even work their way up over damp rocks as Welsh saw them doing in Little River, where they were so plentiful on May 5 and 7, 1913, that he caught 1,500 " The life history of the eel is presented in more detail than is possible here by Schmidt (Philosophical Transactions, Royal Society of London, Series B, Vol. CCXI, 1922 (1923), pp. 179-208, summarized in Nature, Vol. CX, July-December, 1922, p. 716), and by Cunningham (Natm-e, Vol. CXIII. January-June, 1924, p. 199). See also Schmidt (Rapports et Proces-Verbau.x, Conseil Permanent International pour I'E-xploration de la Mer, Vol. V, No. 4, 1906, pp. 137-204. pis. 7-13), and tor a popular exposi- tion of the subject, Smith (National Geographic Magazine, Vol. XXIV, No. 10, October, 1913, p. 1140). FISHES OF THE GULF OF MAINE 81 in one scoop of a small dip net and 545 with a few "grabs" of his bare hand. Elvers in equal multitudes have often been described in other streams — American as well as European. Eels can live out of water so long as to give rise to the story that they often travel overland, but there is no positive evidence for tliis, which leaves their presence in certain ponds having neither outlet nor inlet (a fact often attested) something of a mystery. In a general way it is true, as has so commonly been said, that eels seek muddy bottom and still water, but this is not always so, as large ones are only too common in swift-flowing, sandy trout streams on Cape Cod. The fact is that they can live and thrive wherever food is to be had, and they are at home in high tempera- tures as well as in low ones. Occasionally an eel is caught off the open coast, but this is unusual. No fish is more omnivorous than the eel; no animal food, living or dead, is refused, and its diet in any locality depends less on choice than on what is available. Small fish of many varieties, shrimps, crabs, lobsters, and smaller Crustacea, together with refuse of any kind — for they are scavengers — make up the bulk of the diet in salt estuarine and bracldsh water. Being very greedy, any bait will do to catch an eel. As every fisherman knows, they are chiefly nocturnal in habit, usually Ij^ng buried in the mud by day to venture abroad by night, but eels, large and small, are so often seen swimming about and so often bite the hook by day that this can not be laid down as a general rule. Although very rapacious, the eel grows slowly, the winter rings on its scales (these do not appear until it is 3 or 4 years old — one larval year and two or three in fresh water) having shown that in the case of the European species full grown adults are from 5 to 20 years old, depending on food supph^, etc. This is corrobo- rated for the American species by the fact that Dr. H. M. Smith, former Commis- sioner of the United States Bureau of Fisheries, found a female on the way down the Potomac to be in her twelfth year. The smallest mature males are about 11 to 12 inches long; females 18 inches long. TNIien fully grown, the female eels, traveling mostly at night, drop downstream at the approach of sexual maturity, which takes place inthe fall. They and the maturing males that have been living in the river mouths, bays, and estuaries now cease feeding; the color of the back changes from olive to almost black, while the ventral side turns silvery and the eyes of the males grow to t^vice their previous size. Both males and females then move out to sea. It is not until after they reach salt water that the ovaries mature. In fact no perfectly ripe female eel and only one ripe male (of the European species) has ever been seen. So little is the life history of the eel understood by om* fisher- men that we wish again to emphasize the undoubted fact that no eel ever spawns in fresh water. The eel drops wholly out of sight when once it leaves the shore ; ''° no one knows how deep it swims — whether singly or in companies — but it certainly "Large eels, on their seaward journey, have occasionally been caught by otter trawlers in the western part of the British channel, but we know of no such occurrence on this side of the Atlantic. 82 BULLETIN OF THE BUREAU OF FISHERIES joiirneys out beyond the continental slope into the oceanic basin before depositing its eggs. From the captures of its youngest larvae Schmidt has been able to outline the chief spawning center of the American species as from latitudes 20° to 30° N. and from longitudes 60° to 78° W. — that is, east of Florida and the Bahamas and south of Bermuda, whUe it may spawn (always in deep water) much farther north. Our eel spawns in midwinter, thus occupying one to two months in its journey from the coast to the spawning ground, for Schmidt found very young larvae (7 to 8 mm.) in February. Eels, like Pacific salmon, die after spawning, the evidence of this being that no spent eels have ever been seen and that large eels have never been known to run upstream again. Smith suggests that they probably jellify and disintegrate, as does the conger. Eel eggs have not been seen, but certainly they are provided with an oU globule, as this is present both in unripe ovarian eggs and in the vestiges of the yolk sac of the youngest embryos. Eels (European) are among the most prolific of fish, ordinary females averaging five to ten miUion eggs and the largest ones certainly fifteen to twenty million. The larval, so-caUed "lepto- cephalus" stage (figs. 336, c, and d), as with aU eels, is very different in appearance from the adult, being ribbonlike and perfectly transparent, with small pointed head and very large teeth, though it is generally believed that it takes no food until the time of metamorphosis. These leptocephali, which live near the surface, have been found off oiu- coasts as far north as the Grand Banks, but never east of longitude 50°. Inasmuch as the breeding areas of the American and European eels overlap, not the least interesting phase of the lives of the two is that the larvae of the American species should work to the western side of the Atlantic and the Eiuropean to the eastern side, and that no specimen of the former has ever been taken in Europe or of the latter in America. The American species takes only about a third as long as the European to pass through the larval stage, that is, hardly a year as against two to three years. The leptocephali reach their full length of 60 to 65 mm. by December or January, when metamorphosis to the "elver" takes place, in which the most obvious changes are a shrinkage in the depth and length of the body but an increase in thickness to cylindrical form, loss of the larval teeth, and total alteration in the aspect of head and jaws, while the digestive tract becomes functional. It is not until they approach our shores, however, that the adult pigmentation develops or that the elver begins to feed, a change that is accompanied by a second decrease in size. How such feeble swimmers as the leptocephali find their way in to the neighborhood of the land remains a mystery. It seems certain, however, that all the young eels bound for the Gulf of Maine complete the major part of their metamorphosis while stiU far offshore, not only because we have never taken a leptocephalus in the Gulf of Maine in all our tow-nettings, but (and this is more significant) because the Albatross towed three young eels in the so-called "glass eel" stage, 54 to 59 mm. long — that is, of practically adult form but still transparent — during her spring cruise in 1920, one of them on Georges Bank, March 11; a second on Browns Bank, April 16; and one in the western basin of the Gulf off Cape iVnn, February 23. Evidently they FISHES OF THE GULF OF MAINE 83 were intercepted on their way in to land, and since all three were on the surface we may take it that "glass eels," like leptocephali, keep to the uppermost water layers during their journey. Schmidt has suggested that the American eel is not as plentiful in actual numbers as the European, arguing from the facts that its larvse have not proven so common on the high seas and that the American catch of eels (about 2,000 tons) is but a fraction as large as the European catch (10,000 tons annually). It is not safe to draw any conclusions from the statistics, as the jimerican catch is limited more by the fact that eels are not much in demand than by the available stock. Eels are regularly caught in numbers in muddy bays and in estuaries at the mouths of rivers all along the shores of the Gulf, the catch for 1919 being as follows: Pounds Nova Scotia shore 16, 700 New Brunswick 8, 000 Coast of Maine 305,050 Coast of New Hampshire 2, 000 Massachusetts (including south shore of Cape Cod to Buzzards Bay) 239, 991 This suggests a total of about 400,000 pounds for the Gulf of Maine. The greater part o£ the catch is made by nets and eelpots, with spears a close second, eel spear- ing being carried on chiefly in late autumn and winter in tidal creeks and marshes. v^ -^'Tfi^ ;>- Fig. 34 — Slime eel iSivicnchelys parasiticus) 33. Slime eel {Simenchelys parasiticus Gill) Snub-nosed eel Jordan and Evermann, 1S96-1900, p. 349. Description. — The most diagnostic charactei's of the slime eel — its eel-like form, snub nose, long dorsal fin, and sliminess — have already been mentioned (p. 78). It is stouter and more sway-bellied than the common eel, very soft, and with a more tapering tail. The dorsal fin originates a very short distance behind the tip of the pectoral when the latter is laid back against the body, and the anal runs forward on the lower surface almost to the vent, which is situated about midway of the body. The head is much shorter than in either e,el or conger; the mouth very small, gaping back 84 BULLETIN OF THE BUREAU OF FISHEEIES only about half way to the foi-ward edge of the eye, with upper and lower jaws of equal length and each armed with a single series of small, close-set cutting teeth. The gill openings are very small, and instead of being transverse and on the sides of the neck as in the common eel they are longitudinal and lower down on the surface of the throat. Size. — About 2 feet long. Color. — Dark brown, with belly but little paler than back, though usually more or less silvery. General range. — The continental slope and slopes of the offshore banks, from abreast of the eastern end of Long Island to the Newfoundland Banks, in depths ranging from 200 to more than 900 fathoms; also in deep water about the Azores. Occurrence in the Gulf of Maine. — So far as we can learn there is no definite record of the snub-nosed eel actually within the southern rim of the Gulf. Oiu* only first-hand experience with it was on the slope south of Nantucket lightship in July, 1908, where we captured 21 in the Monaco deep-sea trap in 455 fathoms. It must be extremely abundant along that zone, however, for so many to find their way into the trap in as short a set as two hours. It has been recorded so often in water as shoal as 200 fathoms that it may be expected in the bottom of the Eastern Channel and in the southeastern deeps of the Gulf of Maine. ■ Fig. 35. — Long-nosed eel (Synaphobranchjts f.innatus) Habits. — It is partly parasitic in habit, burrowing into the bodies of halibut and other large fish, under which circumstances a considerable number of specimens have been brought in by fishermen. Very likely it was common inshore in old days when halibut were plentiful there. It also lives independently on the bottom. Beyond this little is known of its manner of life and nothing of its breeding habits. We may add from experience that it is as slimy as a hag — dripping with sheets of mucus when draAvn out of the water. 34. Long-nosed eel (Synaphobranchus pinnatus Gronow) Jordan and Evermann, 1896-1900, p. 351. Description. — This deep-sea species is a typical eel in general appearance but is readily identifiable by the facts that while its dorsal fin originates about as far back, in relation to the length of the fish, as in the common eel (p. 78), its point of origin is behind the vent instead of in front of it, and that the anal fin originates considerably in front of the dorsal fin instead of behind it, as is the case in all other Gulf of Maine eels. Furthermore, its mouth is much wider, gaping far back of the eye, and its snout is pointed. The most interesting anatomic characteristic of this PISHES OF THE GULF OF MAINE 85 eel is that its gill openings, which open longitudinally on the lower side of the throat (in the common eel and in the conger they are situated on the sides of the neck) , join together at the front end to make what apparently is a single V-shaped aperture, though actually they are separate within. Color. — Described as uniform brown, darkest below, with the vertical fins darker behind, pale-edged in front, and the inside of the mouth blue black. Size. — -The largest of 89 specimens measured by Goode and Bean (1883, p. 187) was nearly 22 inches (545 mm.) long, the smallest about 9 inches (221 mm.) in length. CoUett °' mentions one 26 inches (675 mm.) in length from the Azores. General range. — -This deep-water species is of very wide distribution, having been taken near Madeira, off Brazil, off Morocco, near the Cape Verdes, about the Azores, and at many localities off the east coast of North America from the Grand Banks of Newfoundland on the north to the latitude of South Carolina; likewise about the Philippines and in Japanese waters. Most of the captures have been from depths of 300 to 1,000 fathoms. Occurrence in the Gulf of Maine. — This eel has not actually been reported within the geographic limits of the Gulf but is to be expected in the eastern channel and possibly above 150 fathoms along the slopes of Georges Bank, for fishermen have caught them in water as shallow as this off La Have Bank on the one hand, while, on the other, specimens have been trawled in 168 and 129 fathoms off southern New England by the Fish Hawk and Albatross. So many of them have been brought in by fishermen °' from deep water off the fishing banks to the eastward of longitude 65°, and so many have been trawled along the continental slope thence west- ward '"j that this eel must be one of the commonest of fishes below 150 to 200 fathoms from the Grand Banks to abreast of New York. Habits. — Nothing is known of its habits except that it is a ground fish, that the readiness with which it bites proves it predaceous, and that specimens in spawning condition have been taken in summer.'" In its development this species passes through a "leptocephalus" stage even more slender than that of the conger (p. 88)."" •' Rfisultats des Campagnes Scientiflques du Prince de Monaco, Part 10, 1896, p. 164. Monaco. *B Many such instances are listed in the Report of the United States Commissioner of Fish and Fisheries for 1879 (18S2), p. 787 •> Goode and Bean, 1883, p. 187. ™ The "leptocephalus" larvae of the long-nosed eel are described and figured by Schmidt (Rapports et Proces-Verbaui, Conseil Permanent International pour I'Exploration de la Mer, Vol. V, No. 4, 1906, PI. IX, figs. 4-6; and Meddelelser fra Kommissionen for Havunders0gelser, Vol. Ill, Part 1, 1913, p. 14, pi. 2, figs. 1-4. 86 BULLETIN OP THE BUREAU OP PISHEEIES 35. Conger (Leptocephaliis conger Linnseus) Sea EEL Jordan and Evermann, 1896-1900, p. 354. Description. — The readiest characters by which to distinguish the conger from other eels are noted above — notably the origin of the dorsal fin above or only very slightly behind the tip of the pectoral when the latter is laid back, the rather long- pointed snout, the large mouth cleft back at least as far as the middle of the eye, and the scaleless sldn. There are also skeletal differences between this and the common eel," and it has many more vertebras. The conformation of the tip of the snout ■f^W:': ';\;_4^jj^,;jy^^ ^> v ^^ Fia. 36.— Conger eel (Leptocephalus conger) a, Adult, b, Egg. c, Larva, 9 millimeters, d. Larva, 10.2 millimeters, e, Leptocephalus stage, 142 millimeters. After Schmidt. likewise serves to identify the conger, for its upper jaw usually projects beyond the lower, whereas in the eel the reverse is true, or at least the lower equals the upper. Furthermore, the eyes of the conger are oval and larger than the round eyes of the common eel. We need only add further, to give an idea of the proportions of the conger, that the distance from tip of snout to dorsal fin is about one-fifth of the total length, the length of the snout one-fourth that of the head, length of pectorals one-thii'd to one-fourth of the distance from the dorsal fin to tip of snout, and that the body is of the snake-like form characteristic of eels in general. '' For an account of these see Smitt (Scandinavian Fishes, 1892). FISHES OF THE GULF OF MAINE 87 Size. — The female conger grows to an enormous size, certainly to a length of 4}4 to 8 feet, the heaviest creditably reported" weighing 128 pounds (58 kilos). In European seas 50 to 60 pounders are not imusual, and 25 to 30 pounders are usual, but few of this size are caught off New England. Males are very much smaller, perhaps never more than 2J4 feet long. Congers taken near Block Island range from 4 to 6 feet in length; the larger ones taken at Woods Hole usually weigh 8 to 12 pounds. Color. — Grayish-brown above, sometimes of a reddish tinge, sometimes so dark as to be almost black; paler on the sides; dingy white below. General range. — The conger is cosmopolitan in the warmer parts of the Atlantic where it is known as far north as Scandinavia on the eastern side and Cape Cod on the western side. It also occiu-s in the western Pacific and Indian oceans. Occurrence in the Oulf of Maine. — Tlie Gulf of Maine lies north of the regular range of the conger, which is extremely rare there. The curious bandlike "lep- tocephalus" larva of the conger has been taken at Cherrjrfield and at Old Orchard in Maine, and at Nahant in Massachusetts Bay — a total, however, of only six specimens, all of which were collected more than a half century ago — and A. H. Clark, of the United States National Museum, informs us that he found large numbers of leptocephalus larvte, perhaps belonging to the conger, on the beach at Man- chester, Mass., in the summer of 1898 or 1899. The only other positive record for this species in the GuK of Maine is for North Truro on Cape Cod — whether adult or larva was not stated — nor have we ourselves ever seen or heard of an adult conger north of the cape. It is more plentiful west and south of Cape Cod, being taken at Woods Hole from July into the autmnn, and common about Block Island from August until November.'* Habits. — Although the conger is rare in the Gulf of Maine its breeding habits are sufficiently remarkable to deserve brief mention. It is now well established that this species, like the common eel, breeds but once during its life and then perishes. Ripe congers are never caught on hook and line, for they cease to feed — hence to bite — for some time previous, but males kept in aquaria have repeatedly been known to become fully ripe, females nearly so," and then invariably dying, the ripening of the sexual products being accompanied by changes in the shape of the head, loss of the teeth, and a jollification of the bones, while the eyes of the males become enor- mous and the females are much distended by the ovaries. It is probable that the conger ripens off the coast of New England in summer; in captivity they have been kno^\^l to do so every month in the year except October and November. The conger, like the common eel, moves out from the coast to spawn, for its young larvae have never been taken inshore, but if the eggs described below actually belonged to this species, as is generally accepted, then the New England stock travels out only to the edge of the continental shelf for the pm-pose and does not fare forth to far distant parts of the Atlantic Basin as does the common eel. '' Day: The Fishes of Great Britain and Ireland, Text and Atlas, 1880. London and Edinborough. " Local reports of congers do not necessarily relate to the true conger, for the eel pout (p. 378), which is fairly common in the Gulf, is often misnamed thus. " Cunningham (Journal, Marine Biological Association of the United Kingdom, Vol. U, new series, 1891-92, pp. 16-42) gives an interesting account of this and other phases of the life history of the conger. 88 BULLETIN OF THE BUREAU OP FISHERIES The conger is an extremely prolific fish, the number of eggs a female may produce having been estimated as high as from 3 to 6 millions. Conger eggs have never been identified with absolute certainty, but a considerable nimiber of large pelagic eggs towed by the Grampus (presumably on the surface) over the tilefish groimds 30 miles south of Nantucket lightship on July 31, 1900, probably belonged to this species, the larvoB hatched from them undoubtedly being eels while the eggs them- selves were as certainly not those of the common eel. These eggs (fig. 36b) were 2.4 to 2.75 mm. in diameter with 1 to 6 oil globules, one invariably much larger than the others. Tliey hatched in from two to three days in the Woods Hole hatchery, suggesting a total incubation period of four to five days at the prevailing summer temperature. It has long been knoAvn that the conger, like the common eel, passes through a peculiar ribbonlike larval stage — the so-called " leptocephalus " stage — very broad and thin and perfectly transparent, with a very small head. In fact the first "leptocephalus" ever seen (about 1763) was the larval conger, but although its true identity was suspected it was not until 1886, when the famous French zoologist, Delage," actually reared one through its metamorphosis at the biological station at Roscoff, that the identity of this larva was definitely established. The lep- tocephalus of the conger is relatively more slender than that of the common eel, and it can always be identified (imder a lens) by the fact that its vertebra3 and muscle segments are far more numerous (153 to 159 or more, as against only about 107 in the American eel and about 114 in the European eel), and that they grow to a length of 150 to 160 mm. The duration of the larval period of the conger is not known. The process of metamorphosis consists essentially in a thickening and narrowing of the body, an enlargement of the head, the formation of the swun bladder and the permanent teeth, and the pigmentation of the skin, a change that occupied about two months (May to July) in the case of Delage's specimen. At its completion his young conger was 9.3 centimeters (3.6 inches) long." 36. Snipe eel {Nemichihys scolopaceus Richardson) Jordan and Evermann, 1896-1900, p. 369. Description. — ^The snipe eel is made easily recognizable by its extremely slender body (the fish may be 75 times as long as deep), with its tail tapering to a thread, and by its bill-like jaws, which are equall}' elongate, the upper one curving upward but the lower nearly straight. The head is much deeper than the neck, with a large eye. The dorsal fin originates in front of the pectoral, the anal about abreast of the tip of the latter, and both run back to the tip of the tail. There has been some confusion in the published accounts and illustrations as to these two fins, for while Vaillant " shows both about as Ifigh throughout their length as the " Coraptes Rendus hebdomadaires des s&nces de I'Academie des Sciences, vol. 103, 1886, p. 698. Paris. " Schmidtlein (Mittheilungen aus der Zoologischen Station zu Neapel, Band I, 1879, p. 135) speaks of young "congers" at Naples in April as hardly one-third as long as this, a discrepancy suggesting that they may actually have belonged to one of the Muraenoid eels. " Poissons. Expeditions Scientiflques du Travailleur et du Talisman, Pendant les Annte 1S80, 1881, 1882, 1883 (1888), PI. Vir, figs. 2 and 2a. Paris. FISHES OF THE GULF OF MAINE 89 fish is deep, Goode and Bean (1896, pi. 46, fig. 170) picture the dorsal as much higher than the anal (the artist evidently having transposed the two fins) , whereas Brauer '° represents the anal as approximately twice as high as the dorsal and the latter as soft rayed in its anterior and posterior portions but composed of short thornlike spines over its central third. In two specimens taken off New England and now in the collections of the Museum of Comparative Zoology the fins are as follows : Dorsal, soft rayed and nearly as high as the body is deep for its first half; back of that it consists of a series of very short, stiff jays that extend to the tip of the tail. Anal, soft-rayed throughout its length and about as high as the body is deep, tapering to almost nothing on the tail. In part the confusion has been due to the rather fragmentary state in which these deep-water fish usually arrive on board, but at the same time it is probable that two distinct species have been confused under the name scotopaceus, as Brauer Fig. 37. — Snipe eel (Nemicbthys scolopaceus) suspected. Both jaws and the roof of the mouth as well are thickly studded with small sharp teeth. Color. — -Described as pale to dark brown above with the belly and anal fin blackish after preservation. Judging from experience with other deep-sea fishes and from Brauer's plate (which, however, may be another species), we suspect that in life it is chocolate brown above and velvety black below. Size. — Maximum length about 3 feet. General range. — The snipe eel has been taken in the South Atlantic, near Ma- deira, off the Cape Verde Islands, off West Africa, in the Pacific north of New Guinea, and in deep water at many stations off the east coast of North America between latitudes 31° and 42°, longitudes 65° and 75°. Occurrence in the Gulf of Maine. — One specimen taken from the stomach of a codfish caught on Georges Bank in 45 fathoms is the only Gulf of Maine record, but several have been taken in depths of from 300 to 2,000 fathoms seaward from the bank. Habits. — Although commonly spoken of as a "deep-sea" fish, this species is undoubtedly an inhabitant of the mid depths, not of the bottom, and judging from the occurrence of other black fishes it probably finds its upper limit at 100 to 200 " Die Tiefsee-Fische. Wissenschattliche Ergebnisse der Deutschen Tietsee-Eipedition, 1898-1899 (1900), Band XV, Teill p. 126, pi. 9, flg. 1. 90 BULLETIN OF THE BUREAU OF FISHERIES fathoms. Nothing further is known of its habits, but Mowbray's '" recent capture near Bermuda of a snipe eel clinging by its jaws to the tail of a large red snapper has suggested the possibility that this is a regular habit of this curious species. THE TARPONS AND HERRINGS. FAMILIES ELOPID^ffi) AND CLUPEIDJE «' The Elopidre are very closely allied to the true herrings (Clupeidae), from M'hich they differ in having an elongate, bony, interjugular plate on the throat between the branches of the lower jaw. There are only about five species, all of them tropical. The herrings (Clupeidce) are soft-finned fishes wholly lacking spines, with one short dorsal fin, deeply forked tails, ventral fins situated on the abdomen far behind the pectorals, small teeth, deep compressed bodies, and large scales that slip off at a touch. They are, perhaps, the most familiar of northern sea fishes and certainly are the most abundant in number of individuals. Seven spe- cies of herring occur in the Gulf of Maine — the round herring and hickorj' shad, very rare; the sea herring, alewlfe, blueback, and shad, regular; and the men- haden, irregular in its occurrence. Among these the round herring, shad, men- haden, and sea herring are named at a glance, but the others resemble one another so closely that they are often confused even by the fishermen who constantly handle them. KEY TO GULF OF MAINE TARPONS AND HERRINGS 1. Last dorsal ray not elongate 2 Last dorsal ray elongate Tarpon, p. 91 2. Belly rounded Round herring, p. 91 Belly compressed, its edge sharp 3 3. Head (tip of snout to edge of gill cover) not more than one-fourth the total length of the fish; free margins of scales rounded 4 • Head very large, occupying nearly one-third the total length of the fish; free margins of scales fluted instead of rounded Menhaden, p. 118 4. Tip of upper jaw deeply notched, inclosing tip of lower when mouth is closed Shad, p. 113 Tip of the upper jaw is not notched, or, if notched, the lower jaw projects considerably beyond the upper when the mouth is closed 5 5. Edge of the belly is hardly serrated, though sharp; the general form is comparatively shal- low; there is a cluster of teeth on the midline of the roof of the mouth. Sea herring, p. 92 Edge of belly is strongly serrated; general form deep; there are no teeth in the midline of the roof of the mouth 6 6. The lining of the belly is pale gray 7 Lining of belly is black or dark blotched Blueback, p. 110 7. Head occupies about one-fourth of total length of fish; lower jaw projects consider- ably beyond upper; general profile tapers toward the rather pointed head as well as the tail Hickory shad, p. 105 Head occupies hardly one-fifth of total length of fish; lower jaw hardly projects beyond upper; body deep and heavy forward Alewife, p. 107 " Copeia, No. 108, July, 1922, p. 49. " The ladyfish (Elops saurus Linnaeus), a tropical species that would reach the Gulf of Maine as a stray only, has been re- ported by Halkett (1913, p. 45) at Black's Harbor, Charlotte County, New Brunswick, on the north shore of the Bay of Fundy, but the specimen in question, which was 7 inches long and not in very good condition, differed in some respects from published descriptions of this species, hence its identity is doubtful. PISHES OF THE GULF OP MAINE 91 37. Tarpon {Tarpon ailanticus Cuvier and Valenciennes) Jordan and Evermann, 1896-1900, p. 409, fig. 177. Description. — The tarpon is herringlike in general form and appearance, but it is made easily recognizable by the fact that the last ray of the dorsal fin is greatly elongated, its free portion being as long as or longer than the fin is high, and by the presence of the bony interjugular plate mentioned above in the characterization of the family to which it belongs. Furthermore, the anal fin of the tarpon is deeply falcate; that of all Gulf of Maine herrings rhomboid in outline. The ventral fins, which are situated under or behind the dorsal fin in herrings, alewives, shad, and menhaden, are considerably in front of the dorsal fin in the tarpon, while the lower jaw of the latter projects relatively further, its scales are relatively lai^er, and its caudal fin is relatively wider. Fig. 38.— Tarpon ( Tarpon ailanticus) Color. — Bright silvery all over, the back darker than the belly. Size. — Tarpon grow to a length of 6 to S feet (longest recorded, 8 feet 2 inches). General range. — Tropical and subtropical coasts of America, from Brazil to Long Island, casually to Cape Cod, and to Nova Scotia, where it has been recorded twice — off Isaacs Harbor and in Harrigan Cove.*^ Its chief center of abundance is in the West Indies, about Florida, and in the Gulf of Mexico. Occurrence in the Gulf of Maine. — ^A specimen 5J^ feet long, taken at Province- town on July 25, 1915,'^ is the only record of the tarpon in the Gulf of Maine, which it reaches only as an accidental straggler from the south. 38. Round herring {Etrumeus teres DeKay) Stradine Jordan and Evermann, 1896-1900, p. 420. Description. — The most diagnostic feature of this fish among herrings is that its belly is rounded and not sharp edged. It is, furthermore, the most elongate of our hen-ings, its body being only one-sixth as deep as long, thus suggesting a smelt in its general outline. Its dorsal fin, too, stands wholly in front of the ventrals instead >' Halkett, 1913, p. 45. " Radcliffe, 1916, p. 3. 92 BULLETIN OF THE BUREAU OF FISHERIES of over the latter, as in herring, alewives, and shad, and there are fewer anal fin raj'S (only about 13, whereas the herring has about 17, the alewife about 19, and the shad about 21) than any of the latter. Fig. 39. — Round herring {Etrumeus tere Color. — ^Olive green above with silvery sides and belly. Size. — Eight to ten inches. General range. — Atlantic and Gulf of Mexico coasts of the United States; occasionally common as far north as Woods Hole, but very rarely straying past Cape Cod. Occurrence in the Gulf of Maine. — The only published Gulf of Maine records for this southern fish are from Eastport, where the newspapers reported it in 1908, and from Jonesport, Me., but the collection of the Museum of Comparative Zoology also contains two specimens from Provincetown. 39. YL^rving {Clupea harengus 'Lma.siViS) Labrador herring; English herring; Sea herring; Sardine; Sperling; Brit Jordan and Evermann, 1896-1900, p. 421. Description. — This herring is typical of its family in form, with bo^ly so com- pressed that it is much deeper than thick, moderately pointed nose, large mouth situated at the tip of the snout, and lower jaw projecting beyond the upper, which is not notched at the tip (in shad it is so notched, p. 113), sharp-edged belly,?and deeply forked tail. The dorsal fin stands over the small ventrals, its origin about Fig. 40.— Herring (Clupea harengns) midway the length of the body. The scales are large, their rear margins "roimded and so loosely attached as to slip off at a touch. There is no adipose fin, its'absence at once distinguishing herrings from any of the salmon tribe. The dorsal and anal FISHES OF THE GULF OF MAINE 93 fins have no true spines nor is there a lateral line. The anatomical character separating the sea herring from the several alewives (genus Pomolobus) is the presence of an oval patch of small teeth on the vomer bone in the center of the roof of the mouth in the former, but even without this test a practiced eye can separate herrings from alewives at a glance by the comparatively shallow bodies of the former and the deeper outlines of the latter, a difference better shown in the illustra- tions. Furthermore, the sharp midline of the belly is but weakly serrate in the herring, while in the alewives it is saw-toothed. Size. — Herring grow to a length of about 17 inches. Color. — Deep steel blue or greenish blue on the back with green reflections; the sides and belly silvery; the change from dark belly to pale sides often marked by a greenish band. The gill covers sometimes glisten with a golden or brassy gloss; indeed, fish just out of the water are iridescent all over with different hues of blue, green, and violet, beauties that soon fade, however, leaving only the dark back and silvery sides. The ventral and anal fins are transparent white; the pectorals, how- ever, are dark at the base and along the upper edge; the caudal and dorsal dark grayish or shading into green or blue. General range. — Both sides of the North Atlantic. Off the European coast the herring ranges north to Norway, Spitzbergen, and the White Sea, as well as to Ice- land and Greenland; south to the Straits of Gibraltar. On the American coast it is known as far north as northern Labrador ; regularly and commonly as far south as Block Island, though most abundant north of Cape Cod ; and it is occasionally seen at Cape Hatteras in winter. It is replaced by a close ally (C. pallasii) in the North Pacific. Occurrence in the Gulf of Maine. — The herring is probably the most numerous fish in the Gulf of Maine as it is in the North Atlantic generally. To list the locali- ties where it has been recorded would be to mention every hamlet along the coast from which fishing boats put out, for it is universal at one season or another around the whole coast line of the Gulf as well as offshore. Off most parts of the coast, however, the appearance of schools of herring, large or small, is distinctly a seasonal phenomenon, and inasmuch as its on and off shore migrations are intimately bound up with the phenomena of reproduction and growth a brief consideration of these may precede the more detailed discussion of its occurrence. Breeding habits. — ^o much study has been devoted to the natural history of the herring by European zoologists, as well as by Moore (1898) and Huntsman (1919, and unpublished notes) in our own Gulf, and by Lea ^* in more northern Canadian waters, that we have very good knowledge of its breeding habits in general and of its early growth. This may be briefly summarized as follows: It has long been known that the eggs of the herring sink to the bottom, where, by means of their coating of mucus, they stick in layers or clumps to the sand or clay, seaweeds, stones, or any other objects they chance to settle on. They are often found massed on net warps, anchors, and anchor rodes. The eggs are 1 to 1.4 mm. in diameter, depending on the size of the pai'ent fish and also, perhaps, on the local race of fish involved. Females — again according to their age and size — " Age and growth ot the herrings in Canadian waters. Canadian Fisheries Expedition, 1914-15 (1919) pp. 75-164. 94 BULLETIN OF THE BUREAU OF FISHERIES deposit from 20,000 to upwards of 40,000 eggs, averaging about 30,000. In sexually mature herrings, just before spawning commences, the genital organs are so large as to make up about one-fifth the total weight of the fish. The period of incubation is governed by temperature — European students tell us that it requires as long as 40 days at 38.3° F., 1.5 days at 44° to 46° F., and 11 days at 50° to 51° F., while experiments on the Massachusetts coast by the U. S. Bureau of Fisheries gave 10 to 12 days in the temperature prevailing there in autumn. Ten to fifteen days might be stated'as kn average for the Gulf of Maine. Fig. 41. — Eggs of the herring ( Clupea barengus) attached to seaweed, .\fter Ehrenbaurn Spawning grounds in the Oulfof Maine. — Herring spawn all along the shores of the Gulf of Maine wherever the bottom is suitable, from Grand Manan on the east to Cape Cod on the west, as well as along the western shores of Nova Scotia, as, for instance, about the Trinity Ledges (an important breeding ground), thougli whether as universally there we have not been able to learn. Probably the most productive spawning ground within the limits of the Gulf is at the mouth of the ■ Bay of Fundy, particularly about Grand Manan. At present, according to Hunts- man, herring spawn only locally, e. g., at Minas Basin, in the inner part of the bay. FISHES OF THE GUIiF OF MAINE 95 From there westward the presence of spawning herring has been recorded in Machias Bay, about Jonesport, at Mount Desert, in Frenchmans Bay, among the islands at the moutli of Penobscot Bay (Swans, Isle au Haut, and Matinicus *^), in Casco Bay, and especially about Wood Island, a few miles south of Cape Elizabeth, which has long been known as the resort of tremendous breeding schools. Herring also spawn off' the beaches along the western shore of the Gulf — Ipswich Bay, for example, about Cape Ann, in Massachusetts Bay, about Provincetown, along Cape Cod, in the Woods Hole region, near No Mans Land, and about Block Island, which is the southern breeding limit. Spawning takes place both along the shore line and generally on the various reefs and shoals that lie from 5 to 25 miles off the coast of Maine, a habit betrayed by the eggs " that are found adhering to rodes of vessels and boats engaged in the cod and haddock fisheries."*" Indeed, as Moore suggests, it may well be that a large proportion of the herring of our coasts are hatched on these offshore shoals. We find no definite record of herring spawning on Georges or Browns Banks, nor is it likely that they do so on the muddy bottoms of the deeper basins of the Gulf. Herring spawn chiefly on hard, rocky, pebbly, gravelly, or sandy bottoms, to some extent on clay, and probably never on soft mud. Depth of spawning. — Herring are not loio'iMi to spawn in the httoral zone in our Gulf west of Grand Manan (possibly the spring-spawming fish may have done so of old in the Bay of Fundy), nor is the spawn ever cast up on the New England beaches by the surf, a fate that often overtakes it in the Gulf of St. Lawrence. Both in the Bay of Fundy and along the coasts of Maine and Massachusetts eggs are deposited cliieflr from a depth of 2 or 3 down to 30 fathoms. Wliile no absolute depth limits can be established it is not likeh' that our Gulf of Maine herring ever spawn below 75 fathoms (in Scandinavian waters herring occasionally spawn down to even 100 fathoms), for to do so would involve the deposition of the eggs on soft mud bottom, where they would be in danger of smothering. Season of spawning. — It has long been known that, according to locality, the herring spa^vn in spring, summer, or autumn, or in both spring and autumn, as, for example, in the Gulf of St. Lawrence, and it is established that both spring-spawning and summer-fall spawning schools of herring formerly existed in the Baj' of Fundy. These spring spawners visited both the south (Nova Scotian) side of the bay from Brier Island at its mouth in as far as Digby Gut, and the St. Andrews region on the New Brunswick shore near the mouth of the bay, spawning there during April and May. However, they were never very numerous except in restricted localities and have now vanished, temporarily at least. Spring-spawning as well as autumn- spawning herring were likewise reported to us by the fishermen along the west coast of Nova Scotia, though this we have not been able to verify. Other than tliis, spring spawners are neither recorded nor rumored anywhere in the Gulf of Maine. According to Moore the breeding schools arrive in June at Grand Manan, which may fairly be termed the premier spawning ground, to spawn from then until 8« According to fishermen's reports, says Moore, spawning herring were unknown at Matinicus until 18S0; since then the neighborhood has been a productive spawning ground. "• Moore (1898, p. 40-3). 102274— 25t 7 96 BULLETIN OF THE BUREAU OF FISHEKIES late in the fall, with the heaviest runs in July, August, and the first part of September, though Doctor Huntsman writes that in 1917 the season lasted only from early August to early October. In Machias Bay, near by, hei'ring spawn at about the same season, that is, from mid-July until the end of September. Passing thence westward we find the breeding period progressively later and shorter — mid-August imtil October about Petit Manan and near Mount Desert; mid-September until the end of October near Casco Bay and off Wood Island ; September 20 until about Novem- ber 1 in Ipswich Bay;^' the month of October in Massachusetts Bay; while west of Fig. 42. — Larval stages of the herring (Clupea harengas). After Ehrenbaum a. Newly hatched, 7 millimeters. 6. 10 millimeters, c. 19 millimeters, d. 29 millimeters, e. Fry, 41 millimeters. Cape Cod the herring do not begin to spawn until mid or late October, with the major production of eggs about the 1st of November. Thus, while spawTiing occupies three months at the mouth of the Bay of Fundy it lasts hardly longer than one month in the southwestern part of the GuK besides commencing some three months later in the j'ear. Temperature at which spawning taTces place. — Thanks to the considerable num- bers of serial observations taken in the Gulf by the Grampus during the past 10 years and in the Bay of Fundy hj the Biological Board of Canada, it is now possible to estabhsh the temperature at which herring spawn in our waters more closely than « Allen. Memoirs, Boston Society of Natural History, vol. 8, No. 2, 1916, p. 201, jUe E. R Haskell, of Ipswich, Mass. FISHES OF THE GULF OF MAINE 97 Moore was able to do it. Doctor Huntsman has informed us that on the Grand Manan ground (where the temperature is kept practicallj^ uniform, verticalh', b}' the violent tides) herring begin spawning when the water warms to about 46° to 47° F. The temperature there rises to only about 48° at 5 to 20 fathoms depth at the height of the breeding season. Near Eastport, too, the average range of tem- perature from the surface down to 5 fathoms is only about 44.6° to 52° for July, August, and the first half of September, while a comparison of our temperature records with the breeding season at different locaUties shows that in the northern part of the Gulf practically all spawning is carried out in temperatures of 46° to 52°. In the southern part of Massachusetts Bay and along the shores of Cape Cod, M'here autumnal cooling of the surface waters is not as rapid as farther north, herring may spawn in slightly warmer water, say up to 53° or 55°. The Gulf of Maine herring spaA\'n in rather low salinities (such, indeed, characterize the coastal zone as a whole as compared with the North and Norwegian Seas) , the most saline water in which it is known to spa-\vn witliin our limits being not salter than 33 per mille, the freshest probably about 31.9 per mille. They never spawn in brackish water ^ntllin the limits of the Gulf, although known to do so in the almost fresh water of the mouths of certain European rivers. LarvEe of the herring family are extremely slender and can easily be distin- guished from all other young Gulf of Maine fish of similar form (e. g., launce, smelt, or rock eel) by the location of the vent, which is so far back that it lies close to the base of the tail, but it requires critical examination to distinguish our several clupeoids from one another in their early stages. The sea herring is about 5 to 6 mm. long at hatching, with a small yolk sac that is absorbed by the time a length of about 10 mm. is reached. The dorsal fin is formed at 15 to 17 mm.; the anal at about 30 mm.; the ventrals are visible and tail well forked at 20 to 35 mm.; and at about 40 mm. the little fish begins to look like a herring. According to Huntsman's observations fry produced on the Grand Manan spawning grounds in late summer and early autumn grow to a length of 17 to 20 mm. by the end of November or first of December, but there is every reason to believe that the rate of growth slackens during the winter season, not only from studies made elsewhere but because we have taken fry only 26 to 50 mm. long in March and April (p. 100), while young herring 50 to 60 mm. long are abundant in the St. Andrews region in June. These yearlings grow to about 90 to 125 mm. at the end of their first year of life, fish of that size, presumably of the previous autumn 's hatch, being abundant in the fall of the year in the Bay of Fundy. This works out at an average growth of about 10 mm. per month for the warm half of the year, which corresponds fairly closely with the rate at Woods Hole where, according to Smith, young herring spawned in October and early November are 7 mm. long in January, 25 to 32 mm. in May, 65 to 76 mm. in August, and 76 to 125 mm. in their first autumn. In Norwegian waters, too, according to Hjort, the herring is about 125 mm. long at the end of its first year of life,^^ and North Sea herring are 88 Huntsman (1919) believed he could recognize spring as well as autumn-spawned herring fry in the Bay of Fundy, and credits them with a length of about 90 mm. by the first and 150 mm. by the second winter. This, however, would seem to call or confirmation, it being unlikely that herring now spawn in the Bay in spring (p. 95). 98 BULLETIN OF THE BUREAU OF FISHERIES about 100 inin. long when 1 year old. On the average, then, a yearling is between S}4 and 5 inches long. Subsequent growth. — The herring has proved a particularly favorable object for ago and growth studies based on the structure of the scales, a method of investi- gation that has led to a great advance in the knowledge of the life histories of many of our important food fishes during the past 20 years So easily are herring scales interpreted, thanks to their distinct summer and winter rings, '" that confidence can be placed in the age-determinations of the many thousands of herrings that have now been examined and in the correlations between age and size resulting therefrom, thit phase being better known for the herring than for any other fish. Without enter- ing into this subject, which would lead us far afield, we may poiat out that herring not only grow at different rates and that the contrast between the rapid growth of sunmaer and the slow growth of the winter is greater or less in different seas, but that in some localities herring grow rapidly when young and slowly thereafter, while in others they may grow slowly at first but sustain a more even growth to old age, a contrast of this sort obtaining between the herrings of the Magdalen Islands and of the west coast of Newfoundland."" The Dogger Bank herring in the North Sea (to mention a couple of European examples only) approximate 4 inches in length at the end of the first year, 83^ to 9 inches at the end of the third, lO]/i uiches at the end of the sixth, and 11J4 to 12 inches at the end of the ninth year of life, though with considerable variation due, no doubt, to varying food supply and to the general suitability of the conditions under which they live. According to Hjort °' Norwegian herring spawned in the year 1899 and examined at 10 years of age exhibited the following growth: 33^ inches at 1 year, 734 inches at 3 yeai-s, 11 34 inches at 6 years, 123^ inches at 9 yeare, and about 12% inches at 10 years. From this it is evident that they did not grow 90 fast as the North Sea fish at first, but attained the same size at 6 to 9 years of age. Huntsman (1919) credits the Bay of Fundy herring with about the same growth — 4 inches at the end of its first and 10 inches at the end of the third year — as the Dogger Bank fish, maldng most of their growth from May to September. Probably the growth period lasts a month later in the southern parts of our Gulf. In Nor- wegian watei's it has been found that herring grow from April to September only, remaining practically stationary in length from October imtil March. "- Size at maturity. — According to Moore, who examined thousands of fish about Eastport, herring rarely spawn when less than 93^ inches long, usually not until they are 10 to 103^ inches, with most of the spa^smers 12 to 13 inches long. This he interpreted to mean that some few spawn when only 2 or 3 years old — most of them, however, first at 4 years or older — to continue spawning annuallj' '» See Einar Lea (Age and growth of the herrings in Canadian waters. Canadian Fisheries Expedition, 1914-15 (1919,) pp. 75-164) for an account of age-determination by analysis of the scales, as it applies to the herring. »» Hjort, Canadian Fisheries Expedition, 1914-15 (1919), pp. xi-xiviii, and Lea, /lid., pp. 75-164. •' Conseil Permanent International pour I'Exploration de la Mer, Rapports et Proces-Verbaux, Vol. XX, 1914, 228 pp. 3 pis. Copenhague. •' Lea, Einar; A study on the growth of herrings. Conseil Permanent Internatonal pour I'Exploration dela Mer, Publica- tions de Circonstance No. 61 (1911), pp. 35-57. Hjort, Johan; Fluctuations in the great fisheries of northern Europe, viewed in the light of biological research. Conseil Permanent International pour I'Exploration de la Mer, Rapports et Proces-Verbaux, Vol. XX, 1914, 228 pp., 3 pls. Copenhague. FISHES OF THE GULF OF MAINE 99 thereafter as long as they live. In Norwegian waters, too, a few spawn at 3, many at 4, and the majority at 5 years of age and upwards; some few, indeed, not until 6 years. Herring as old as 20 years have been seen, and they may live even longer. Success of reproduction. — The maintenance of the stock of any fish and its relative abundance from year to year depends less on how many fish spawn in any locality than on how many of the resultant fry survive. The many age analyses made of herring of different sizes and from various seas have proven that while in some years a very large crop of young fish is produced, in others hardly any are obtained even in favorable nurseries. Apparently this applies more to the northern than to the southern breeding grounds — to some extent, however, to all — the result being that the fish spawned in some one favorable breeding season may dominate the herring schools over large areas for many years or imtil another good breeding year produces another large crop. In Norwegian waters, for example, few herring were raised in 1903 but so many were produced in 1904 that fish of that year pre- dominated in the catches for the next six yeare at least. Unfortunately information along this line is yet unavailable from the Gulf of Maine. °' No doubt similar fluctuations occur in the crop there, too, for Lea"' found that fully 50 per cent of the herring taken at West Ardoise and Lockport on the outer coast of Nova Scotia in 1914 belonged to the year-class spawned in 1911, whereas on the west coast of Newfoundland fish hatched in 1904 dominated the spring catches of 1914 and 1915. Various explanations, such as abundance or scarcity of microscopic plankton, favorable or unfavorable tempera- ture or salinity, etc., have been proposed to account for this, aU of which may enter in, for while it is during the first few weeks of life that the herring is most vulnerable it is also possible that the conditions under which the parent fish lived for the year preceding spawning may influence the fate of the fry. Whatever the explanation, the fact that such fluctuations do occur from year to year in the stock of fry reared is of the greatest practical interest to all concerned with the sea fisheries, as evi- dence that variations existing in the stock of herring, and consequently in the catch, are due more to the success or failm-e of reproduction than to depletion by over fishing. Seasonal movements of herring in the Gulf of Maine. — The life of the herring may be divided, roughly, into three stages correlated with differences in distribu- tion and seasonal movements. First, the young and "sardine"; second, the imma- ture "fat"; and third, the mature "spawn." When the little herring reach an age of about 2 years and a length of 190 to 200 mm. they begin to accumulate large amounts of fat among the body tissues and viscera during their period of active growth in the warm months of the year, and lose this fat in winter and at the approach of sexual matiu-ity. We can bear witness — the fact is well known to fishermen — that the "fat" stage is as characteristic of American waters as of European, where "fat" herring are the objects of extensive fisheries. Owing to the fact that most of the herring larvje hatch and pass the first couple of months of their existence at a time of year (September to February) when we " Herring studies had been one ot Mr. Welsh's major undertaking.^. " Canadian Fisheries Expedition, 1914-15 (1919), p. 131, flg. 38. 100 BULLETIN OP THE BUREAU OF FISHERIES have made few hauls in the coastal waters of their nativity, we have taken very few smaller than 20 mm. long in our tow nets. The largest catch was in Gloucester Harbor where larv£e of only 9 to 11 mm. (spawned but a few weeks previous) swarmed on October 24, 1916. Huntsman, however, has been able to follow the migrations of September-spawned herring off Grand Manan,"^ his general and we believe justifiable conclusion being that for a short time the larva are carried passively in the water, resulting in a drift to the southwest, with the set out of the Bay of Fundy, at a rate of about 2 miles per day, but that they turn back and make head- way against the current (technically become "contranatant") when no more than 18 mm. long.'"' Probably the young herring sink down into the water a few fathoms deep during their fii'st winter to escape the extreme chilling of the surface stratum, while our tow nettings afford evidence that before the following spring they become very widely dispersed over the Gulf, for during March and April of 1920 we took odd specimens at localities as generally distributed as the neighborhood of Cashes Ledge, the northern and eastern parts of Georges Bank, the north Channel, off Seal Island and Yarmouth (Nova Scotia), off Lurcher Shoal, off Machias, Me., both near land and out over the deep basin, near Boothbay, and near the Isles of Shoals. It is probable, however, that the majority of any particular body of fry hatched together remain near their birthplace, for not only may little herring be taken just outside the Bay of Fundy in winter (though they desert its estuaries then), but they reappear, grown to a length of 2 to 3 inches, along our entire shore line in myraids with the advent of spring. This reappearance takes place about the middle of April — sometimes as early as the last of March — in Massachusetts Bay; in April and May along the eastern coast of Maine, in the Bay of Fundy, and on the west coast of Nova Scotia. East of Penobscot Bay generally, and particu- larly about the mouth of the Bay of Fundy, "sardine" size herring from 3 to 8 inches long, including 1 and 2 year olds, maybe expected in abundance all summer, though the schools wander and are so local in their appearances and disappearances that they may swarm in one bay while sought in vain a few miles away. In the southwestern part of the Gulf, however, as exemplified by Massachusetts Bay, it is probable (though not yet proven) that yearlings do not appear until several months later than the fish of 2 to 3 inches hatched the preceding autumn, for while these are reported as more or less abundant throughout the spring," especially in such partially inclosed waters as Provincetown Harbor and Plymouth and Duxbury Bays, they apparently move out again during the early part of summer, being far less plentiful in June than in April and May, and it is not until late July or August that "sperhng" of 5 to 7 inches (fish in then- second summer) appear in numbers off the Massachusetts coast. Even in a region as small as Massachusetts Bay wide local variation obtains in the abundance and time of appearance of the "sperling." At Provincetown. for example, they may be expected in schools plentiful enough to be worth "torching" w Doctor Huntsman allows us to quote from unpublished notes. ■ » This requires confirmation, as Doctor Huntsman remarks in his notes. •' Being too small for bait, and there being no sardine factories on Massachusetts Bay, no attention is paid to the smallest herring there, and consequently little is known about them during their first spring and summer. FISHES OF THE GULF OF MAINE 101 any time after mid-July. At Cohasset, however, near the head of the Bay, this seldom happens before September, but we ourselves have seen "sperling" very plentiful there, singly and in small companies, in July and August. During the autiunn these two year classes (the fish in their first year have grown to a length of 3 or 4 inches; those in their second year to 7 to 9 inches by September) together with the large spa-\vning adults are very abundant all around the shore line of our Gulf, but they begin to thin out after the middle of October, and when mnter sets in the fish that are then 1 and 2 years old move out into deeper water once more, few being taken after early December. Proliabh^ they winter mostly on the bottom, for schools are very seldom reported on the surface then, but there is no reason to suppose that the bulk of them travel far, for herring of all sizes are to be found in the Bay of Fundy all mnter and are even caught occasionally in the weirs near Eastport as late as February, being seen again as early as March and April. It is during the third summer, when the Gulf of Maine herring is past the "sardine" or "sperling" stage and has not yet reached spawning age, that least is known about its movements. It is now "fat" and termed a "sea" or "summer" herring locally. Our "fat" herring lie offshore more than do the younger fish, and although numbers of them are taken in the weirs and traps all along the coast there seems to be no definite run of them inshore. On the other hand they are often met far at sea, and it is generally taken for granted that the schools of herring encountered out in the open Gulf in summer bfelong to this category, for when a mackerel seiner picks up svich a school "^ the fish usually are very fat and show no signs of approaching sexual maturity. Some years' these "summer" herring, weighing about 1 pound and very fat (locally they are called "spawn" herring, but this is an error), are taken'in the traps at Provincetown for a week or so about mid-April. They are met at about the same time off Gloucester (in 1915 they were reported 8 to 15 miles off Cape Ann on the 17th), and they are said by the fishermen to work eastward thereafter, being found off Seguin in May and June and off Mount Desert in late summer. As a rule few of them are taken inside the islands, but these "fat" herring came right into the harbor of Boothbay about May 14 in 1914. Fishermen universallj^ agree that they follow the coast only as far east as Mount Desert Rock, hence it is prob- able that they pass the late summer and early autumn offshore in the northeast corner of the Gulf, after which they drop out of sight. The large mature herring (in the fourth summer and older) live some distance offshore during most of the year and, as European experience suggests, near bottom, coming inshore only to spawn, for they are neither caught along shore nor seen schooling on the surface except for a brief period before, during, and after the spawning season. Since they are to be found throughout the year in the Bay of Fundy, however, it is not probable that they travel far. The date of their appear- ance on the coast depends not only on the date v/hen spa^vning commences (p. 96), but to some extent on purely geographical conditions, for they show about the ss Many events of this sort have been reported. For example, a large catch of fat summer herring was made on Georges Bank and reported to the Massacliusetts Commissioners in the midsummer of 1901. 102 BULLETIN OF THE BUREAU OF FISHERIES off-lying islands some time before they make their way up the bays, two or three weeks earlier, for instance, at Grand Manan, Jonesport, and about Mount Desert Island than within Machias Bay. At Isle au Haut, at the eastern entrance of Penobscot Bay, and at Castine, within tne Bay, they appear some time after the middle of July, but not until the end of that month or the first of August at Matinicus, while tlioy are not to be expected in the Massachusetts Bay region imtU the last week in September. By October they are in full force all along the shore of the Gulf from Cape Cod to Grand Manan, and are equally widespread, if less abundant, inshore in November, occasionally being reported in December and even later. It is probable that as the fish spawn out they move out promptly from the spawning grounds into deep water, for but few recently spent fish are taken in the Weil's. Summary. — The young herring 1 and 2 years of age come inshore early in spring and spend the summer and autumn on the New England, New Brunswick, and Nova Scotian coasts; the sexually mature herring come in onh' to spawn, while the "fat" immature herring as a rule summer in the deeper M'ater outside the outer islands. A question often asked but which can not yet be answered in a satisfactory way is "jiist where do the Gulf of Maine herring winter?" In the case of the spring spawners that formerly inhabited the Bay of Fundy the wintering ground apparently lay between Grand Manan and the neighboring mainland of New Brunswick, where a considerable Avinter fishery was formerly carried on. Probably the autumn spawners, both young and adult, merely descend into deeper water to winter, as is the case in European waters, but how far or how deep the great body of them go is not known. It has been proven, however, that herring of practically all ages remain in the open Bay of Fundy throughout the cold season, and the abundance in the deeper v/ater layers of the northeast corner of the Gulf of pelagic euphausiid shrimps, a favorite herring food, suggests this as a rich winter pasture for them. Herring on the offshore hanlcs. — Very little is known about the status of the herring on the offshore banks. Occasionally schools (invariably proving to be "fat" if captured) are encountered on Georges Bank and in the deep water to the north of it (p. 101). During the early years of the Georges Bank cod fishery (about the middle of the past century) great schools of herring were seen there and the fishermen made a regular practice of setting herring drift nets for bait, but the facts that the beam trawlere very seldom catch herring there (then only an odd fish) and that the stomachs of cod and haddock caught on the Bank seldom if ever contain herring,^' is evidence that no great body ' of the latter seeks the Bank for wintering. Since sperling are unknown on Georges Bank — a fact commented on by Storer long ago — it seems that herring seldom move so far out to sea until they are 2 years old. ^ W. F. Clapp found no herrinp in many cod and haddock stomachs from Georges. ' During the beam trawler investigations of 1913 herring were reported for almost every month in the year, never, however, more than a dozen or so fish on any trip, and usually only one or two. FISHES OF THE GULF OF MAINE 103 Food. — The herring is a plankton feeder. When first hatched, and before the disappearance of the yolk sac, the larva (European) feed on larval gastropods, diatoms, peridinians, and crustacean larvae, but they soon begin taking copepods, and after they are 12 mm. long depend on them exclusively for a time, particidarly onj^the little Pseudocalanus elongatus} As they grow older they feed more and more on larger prey, turning to the larger copepods and amphipods, pelagic shrimps, and decapod crustacean larvse. Examination of 1,500 stomachs' showed that adult herring near Eastport were living solely on copepods and pelagic shrimps, fish less than 4 inches long depending on the former only while the larger herring were eating both. When feeding on copepods herring swim open-mouthed, often with their snouts at the surface, crossing and recrossing in their tracks and evidently straining out the minute crustaceans by means of their branchial sieves, a straining apparatus of coarser mesh than that of the menhaden and consequently capturing larger plankton and letting the microscopic plants pass through. When feeding on euphausiids, as we ourselves have often seen them engaged and with which the large fish are often gorged, they pursue the individual shrimps, which often leap clear of the water in their eft'orts to escape. Even in winter when shrimp are rarely seen on the surface Moore found them an important article in the diet of the herring, and it is not unlikely that the local appearances and disappearances of schools of large fish in the open Gulf are connected with the presence or absence of shrimp. In the Gulf of Maine these pelagic shrimp (euphau- siids) are taken by hen"ing in preference to any other food, and are voluntarily selected from among the hosts of copepods by such fish as are large enough to devour them. Even when both shrimp and copepods abound, however, a few of the larger fish, as well as the smaller, will usually be foimd full of copepods, though most of them are packed with shrimp, and in the absence of shrimp (which are seldom abundant west of Mount Desert except during brief periods) copepods are the chief dependence of all our herring, large and small. Such, for instance, is the case at Woods Hole, where copepods had been the chief diet of almost all the herring examined by Doctor Linton during the summer of 1918, and there can be little doubt that they actually select copepods in preference to other small floating organisms, for they are often found packed with them at times and places when the tow nets reveal the presence of a great variety of other animals. In European seas the amphipod genus Euthemisto is also an important food for herring, hence it is to no hesitancy to capture them that the absence of Euthemisto from the herring stomachs examined by Moore and by us is due, but to the comparative scarcity of this large active crustacean in the coastwise waters of the Gulf of Maine. In defaidt of an abundant supply of Crustacea, and sometimes even when these are plentiful, herring feed on whatever molluscan larvse, fish eggs, Sagittae, pteropods, annelids, etc., the water contains, even on objects as small as tintinnids and Halosphsera, but the smaller microscopic plants, either diatonr or Peridinian, are never found in the stomachs of herring more than 15 to 20 mm. long, probably ' The diet of the young herring in the English Channel has been described by Lebour in a series of paper.s, especially in Journal, Marine Biological Association of the United Kingdom, Vol. XII, September, 1921, pp. 458-16V. Plymouth. J Moore, 1898, p. 402. 102274— liot S 104 BULLETIN OF THE BUREAU OF FISHERIES because their gill rakers are not fine enough to retain them. Although herring normally are not fish eaters, small launce, silversides, and the young of their own species have been found in them at Woods Hole. Unfortunately the particular species of copepods on which Gulf of Maine and Woods Hole herring gorge have not been identified, but we might guess that Calanus, with Pseudocalanus, Acartia, and Centropages predominate, while at its times of abundance Temora no doubt looms large in the diet of the herring here as it does in the Irish Sea, and Euchseta offers a rich food supply when the schools seelv the deep waters of the basin where these mammoth copepods abound. Enemies. — The herring is the best of all bait in the Gulf. Naturally, then, it is preyed upon by all kinds of predaccous fish, especially by cod, pollock, haddock, silver hake, mackerel, salmon, dogfish and other sharks. Silver hake in particular often drive schools of herring right up on our beaches, where pursued and pursuers alike strand on the shoaling bottom. The finback whales also devour them in great quantities, and the common squid (Ommastrephes) destroys multitudes of the young sardines. Destruction hy natural causes. — The herring is a very "tender" fish, prover- bially prone to wholesale destruction by stranding on beaches during storms and by pollution of the water. Many instances of this kind have been reported. Allen,* for example, saw young herring in windrows for miles on the strand at Rye Beach in August, 1911. A slaughter of herring, more instructive because the exact course of events was followed, occurred at Cohasset, on the south shore of Massa- chusetts Bay, in October, 1920. On the .5th of that month a large school of Sper- ling, 4 to 5 inches in length, ran up the harbor (which is nearly landlocked), prob- ably driven in by silver hake (at least so local fishermen said), were trapped there by the falling tide, and stranded on the mud. So numerous were they that the flats were entirely covered with them and it was estimated that 20,000 barrels of fish perished. During the next few days the fish, alternately covered and uncov- ered by the tide, decayed and in spite of tidal circulation so fouled the water that lobsters died in the floating cars. On the 10th there was a second but smaller run of herring, and on the 15th a run as large as the first occurred, the newcomers dying soon after they entered the harbor. Altogether, it was estimated that 50,000 barrels of fish perished, of which over 90 per cent were sperling, 5 to 10 per cent were large adults, and a few were small mackerel and silver hake, besides large numbers of smelt. By the last half of October, when I saw them, the flats were silvery with herring scales at low tide, and the residents about the harbor found the stench almost imbearable. During the winter months the fish entirely decomposed and the water purified itself. In north European waters vast quanti- ties of herring spawn are likewise cast up on the beaches every year to perish. Annual fluctuation in the supply of herring. — Many times during the past 75 years the complaint has been made that the herring of the Gulf of Maine are diminishing in number, but Moor' 898), who sifted many sources of information, concluded (we believe rightly) that t ere had been no general decrease in the abundance < Memoirs, Boston Society of Natural History, vol. 8, No. 2, 1916, p. 202. FISHES OF THE GULF OF MAINE 105 of young herring at the mouth of the Bay of Fundy up to that time, though the numbers of herring visiting any given locaUty on our coast and the duration of their stay vary widely from year to year. As far back as the period from 1837 to 1857 Massachusetts Bay saw a marked diminution of the local supply of herring followed shortly by its reestablishment in fuU strength, and any particidar locality — for instance the Eastport region or Wood Island — would no doubt show similar ups and downs from year to year or over periods of years. Local spawning grounds, too, may be abandoned for a term of years — a very common occurrence.^ The fact that the catch varies widely from year to year is not governed altogether by the abundance of the fish themselves, for sundry economic factors enter in, and except for the disappearance of the spring spawners from the Bay of Fundy no general alteration, one way or another, far-reaching enough to have impressed itself unmistakably on those chiefly concerned, has taken place in the herring supply of the Gulf in recent years. In short, Capt. John Smith's (1616, p. 188) account of the herring applies as well to-day: "The savages compare the store in the sea with the hairs of their heads, and surely there are an incredible abundance upon this coast." In the year 1919 more than 110,000,000 pounds of herring were caught in the Gulf, about 3,400,000 pounds being taken on the Nova Scotia shore, 10,415,000 on the New Brunswick shore, 86,700,000 off the Maine coast, and 10,800,000 off Massachusetts. Since at least 80 per cent of the total catch consisted of "sardines," that is, of fish of only a few ounces weight, the toll taken can not have been less than half a billion fish. Unfortunately, however, present plenty is no guaranty of permanent abundance, for the history of the herring fisheries on the other side of the Atlantic, where the record runs back for centuries, has been a succession of periods of plenty and of scarcity since the earliest times. 40. Hickory shad {Pomolohus mediocris Mitchill) Fall herring ; Shad herring Jordan and Evermann, 1896-1900, p. 425. Description. — The hickory shad is distinguishable from the common sea herring by the absence of vomerine teeth, its very deep belly, upper jaw notched at the tip, and the fact that its outline tapers toward both snout and tail in side view (fig. 43). A hickory shad 13}^ inches long is about 4 inches deep, while a herring of that length is only 3 inches deep. Furthermore, in the hickorj^ shad the lower jaw projects much more than in the herring and the dorsal fin originates nearer the snout than the tail, whereas in the herring it is about midway of the length of the body though the difference in this respect is not great. One is more likely to con- fuse this fish with the alewives, which resemble it in the great depth of the body. It can be distinguished from the blueback by the color of the lining of the beUy, which ' Moore, 1S98, p. 430. 106 BULLETIN OF THE BUREAU OF FISHERIES is pale gray instead of black, and there is also a slight difference in outline, the hickory shad being the deeper of the tM^o, with a more tapering head, its lower jaw projecting relatively farther. The hickory shad is distinguishable from the common alewife (in which the hning of the belly is Hkewise gray) by tapering fonv^ard from its greatest depth, whereas the latter is heavy bodied foi-ward of the doreal and ventral fins, as well as by its sharp head, longer jaw, smaller dorsal fin, only about one-half as many gih rakers (about 30 on the first gill arch), and, under favorable circumstances, by its color, being faintly marked on the sides with longitudinal stripes. The projecting lower jaw of the hickory shad marks it off from tlie shad. Size. — This, next to the shad, is the largest of our anadromous herrings, growing to a length of 2 feet and a weiglit of 23/^ pounds. General range. — Atlantic coast of North America from the Bay of Fundy to Florida, running up into fresh water to spawn. Occurrence in the Gvlfof Maine, — The hickory shad is a more southern fish than either of the alewives, the Gulf of Maine marking the extreme northern limit to its range. So far as known, it does not breed in any of the rivers north of Cape .KSfi^ Fig. 43. — Hickory shad (Fomolobus mediocris) Cod and is so rare a fish within its limits that although recorded from the mouth of the Bay of Fundy (Huntsman doubts this record), from Casco Bay, off Portland, at the mouths of various I'ivers in Maine where odd lish are taken from time to tune by gill-netters, in Boston Harbor, at Provincetown, and at North Truro in Massa- chusetts, we have not seen a single specunen among the thousands of herring and alewives that have passed through our hands. West of Cape Cod, however, it is much more plentiful, being common from spring throughout summer and early autumn at Woods Hole, where as many as 3,500 have been taken at a single lift of one trap. In 1919 the Massachusetts catch of hickory shad, practically all from the .south coast, amounted to 12,800 pounds. Habits. — Nothing is known of the liabits of the hickory shad in the sea to differ- entiate it from its close relatives, the alewives, except that it is more of a fish eater, as might be expected from its large mouth and strong jaws. Launce, anchovies, cunners, herring, scup, silversides, and other small fish, squid, fish eggs, and even small crabs, as well as sundry pelagic Crustacea, have been found in the stomachs of hickory shad at Woods Hole by Vinal Edwards. FISHES OF THE GX7U OF MAINE 107 41. Alewife {Pomolobus pseudoharengus Wilson) Gaspereau; Sawbelly; Kyak; Branch herring; Fresh- water herring; Grayback Jordan and Evermann, 1896-1900, p. 426. Description. — The lack of vomerine teeth distinguishes the alewife, with its brethren, the hickory shad (p. 105) and blueback (p. 110), from the sea herring, but even without the slight examination of the mouth which determination of this point entails, it is distinguishable at a glance from it by the greater depth of its body, which is three and one-third times as long as deep — an alewife of 13^ inches being about 4 inches deep while a herring has a depth of only 3 inches. Furthermore, the alewife is much more heavily built forward than is the herring, and the serrations of the midline of its belly are much stronger and sharper — hence the local name "saw- belly" — so much so that a practiced hand can separate herring from alewives in the dark. The most useful distinction between the alewife and the blueback is the fact Fig. 44. — Alewife (Pomolobiis pseudoha'engvs) that the hning of the belly is pale in the former and black in the latter. Alewives are distinguished from young shad by the projection of the tip of the lower jaw beyond the upper when the mouth is closed, and by the fact that the outline of the edge of the lower jaw is deeply concave in the alewife and nearly straight in the shad. Color.— The alewife, like the herring, is dark green blue above, darkest on the back, paler and silvery on sides and belly. Usually there is a dusky spot on either side just behind the margin of the gill cover (lacking in the herring) and in large fish the upper side may be faintly striped with dark longitudinal lines. In life the sides are iridescent with lines of green and violet. Size. — The alewife grows to a length of about 1 foot, adults averaging about 10 inches long and slightly more than half a pound in weight; 16,400,000 taken in New England in 1898 weighing about 8,800,000 pounds. General range. — -Nova Scotia and the Gulf of St. Lawrence south to the CaroUnas, running up into fresh water to spawn. Occurrence in the Gulf of Maine. — When the white man crossed the Atlantic probably there was no stream from Cape Sable to Cape Cod but saw its annual run 108 BULLETIN OF THE BUREAU OF FISHERIES of alewives unless they were baiTed by impassable falls near the mouth. In the words of an eyewatness, "experience hath taught them at New Plymouth that in April there is a fish much like a herring that comes up into the small brooks to spawn, and when the water is not knee deep they will presse up through your hands, yea, thow you beat at them with cudgels, and in such abundance as is incredible." " During the past two centuries, however, its numbei-s have declined and its range has been restricted, both by actual extirpation from certain streams by over- fishing, by the pollution of the river waters by manufacturing wastes, and by the erection of dams that it can not pass. However, the alewife is still a familiar fish all along our coast, ^ and yields an abundant catch in many of om- streams. Ale- wives are taken commonly about Yarmouth, Nova Scotia, and they are locally abundant in the Bay of Fundy, e. g., in Minas Channel in the Annapolis Basin as well as at other localities farther up the bay, with large runs in the St. John River ; and passing along the coast of Maine we find them entering both the large river systems and their tributaries and innumerable small streams, the one requirement being that these shall lead to ponds. At Boothbay Harbor, for instance, a con- siderable stock of alewives annually runs up to spawn in Campbell's Pond, a small body of water dammed off from the harbor and reached by a short fishway only 15 feet long. Perhaps this is our shortest alewife stream. Alewives also breed in many ponds lying back of barrier beaches, which they enter through artificial cuts opened on purpose. To show how catholic the alewife is in its choice of rivers we may point out that in 1896, when the fishery was the subject of inquuy by the Bureau of Fisheries,^ catches large enough to be worth special notice were reported from the mouths of the St. Croix, Dennys, Machias, Medomak, Penobscot, St. George, Pemaquid, Damariscotta, and Kennebec Rivers, from Casco Bay, and from sundry other shore localities in Maine, from the Pis- cataqua River system in New Hampshire, the mouth of the Merrimac, and from Cape Cod Bay in Massachusetts north of Cape Cod, but few alewives now ascend the Merrimac, so polluted is it and obstructed by dams, though fishways recently constructed now allow some to ascend beyond Lowell, Mass. In 1921 Belding found them still running in only about 12 streams on the Gulf of Maine coast of Massachusetts (and very few in these) out of 27 streams that formerly supported considerable alewife fisheries. The fact that in 1896, 5,832,900 were caught along the coast and river mouths of Maine, 526,500 in New Hampshire, 2,677,972 as the combined catch of the Merrimac River and of Cape Cod Bay, suggesting a total of not less than 3,000,000 for Massachusetts north of Cape Cod,^ i. e., at least 9,300,- 000 (in actuality probably considerably more than 10,000,090) alewives of market- able size from the western and northern shores of the Gulf, will illustrate the numerical strength of this fish. This does not include the yield of tlie Bay of Fundy « Capt. Charles Whitborne. in "The True Travels of Capt. John Smith," etc., 1616, vol. 2, p. 250. ' Belding (1921) has given a very instructive report on the alewife in Massachusetts. » Smith, H. M. Report, U. S. Commissioner of Fish and Fisheries, 1898 (1899), pp. 31-43. • The total alewife catch for Massachusetts was about 10,000.000 flsh, but most of these were from the streams emptying on Ibo other side of Cape Cod. FISHES OF THE GULF OF MAINE 109 nor of the west coast of Nova Scotia, for which no precise statistics are available. Large though these figures are, however, and numerous though the alewives seem when crowding into streams, they make but a sparse population as compared to the sea herring when spread over the Gulf. For example, in the year 1920 seiners and drift-netters fishing offshore brought into Boston, Gloucester, and Portland only 31,650 pounds of alewives and bluebacks combined, as contrasted with 4,000,000 pounds of lierring.'" Habits. — The first alewives appear early in April in the few streams tributary to Massachusetts Bay that they still frequent, but the}'^ are seldom seen in Maine rivers or in the St. John until late April or early May, a difference in date probably depending on the temperature of the water. Thereafter successive runs follow (the last half of May seeing the heaviest) until well into June. In 191.5, for example, alewives appeared in Campbell's Creek at Boothbay Harbor on April 20, and were stiU running as late as May 20, by which date spent fish on their return trip to salt water were passing those coming in. During the early runs sometimes one sex predominates, sometimes the other, but as a rule tlie late runs consist chiefly of males and these greatly outnumber the females on the spawning grounds. The alewife spa^vns in ponds, never in running streamy, each female depositing from 60,000 to 100,000 eggs, according to her size." Spawning lasts only a few days for each group of fish, taking place at a temperature of 55° to 60°, the spent fish running downstream again soon after spawning, some commencing this return journey as early as May. Incubation occupies 6 days at a temperature of 60°. Tlie young alewives,'^ which are about 5 mm. long when hatched, growing to 15 mm. when a month old, soon begin to work their way downstream, successive com- panies of fry moving out of the pond and down with the current throughout the simimer. They have been seen descending as early as June 15 and by autumn, when 2 to 4 inches long, the young alewives have all found their way down to salt water. Thenceforth the alewife lives in the sea until sexually mature and very little is known about its habits or migrations. As every fisherman knows, it is as gregarious as the herring, fish of a size congregating in schools of thousands of individuals (we find record of 40,000 fish caught in one seine haul in Boston Harbor) and apparently any school holds together during most of its sojourn in salt water. At times, however, alewives are caught mixed with menhaden, herring, or blue- backs. Although alewives, immature and adult, are often picked up in abundance in weirs here and there along the coast, it seems that most of them, like the "fat" herring (p. 101), keep outside the islands, and the fact that odd alewives were re- ported from Georges Bank in March, June, and August, 1913 (39, indeed, were taken on one trip) , and a few caught in the trawls in the South Channel in Novem- ber, proves that they may wander far offshore. The alewife, like the herring, drops out of sight in winter, but probably it simply moves offshore then, living near w This takes no account of the tremendous shore catch of herring mentioned on p. 105. " The average number of eggs in 644 females taken in the Potomac was 102,800. (Smith, H. M. North Carolina Qeological and Economic Survey, Vol. II, 1907, pp. 1-449. Raleigh.) " The development of the eggs, larv.il stages, and young fry are described by Ryiler (Report, U. S. Commissioner of Fish and Fisheries, 1885 (1887), p. 505) and by Prince (1907. p. 95). 110 BULLETIN OF THE BUREAU OF FISHEHIES the bottom like the latter, for Huntsman (1922a, p. 10) reports its young at Campobello Island, Bay of Fundy, in December and March. The rate of growth of the alew'ife during its sojourn in the sea has not been studied, but experiments in stocking ponds with alewives long ago led to the con- clusion that they attain sexual maturity at 3 or 4 years of age. This has been confirmed on many occasions in recent years, for it has proved easy to restock suitable streams with alewives by planting adult spawners at the proper season in the ponds which they drain, as has been done in many localities in Massachusetts. In the third year after the "plant," and not until then, tlie progeny appear in the stream. The success of such restocldng operations, moreover, has demonstrated that the "parent stream" theory — that is, that alewiyes, like shad, return to spaA\Ti in the stream in which they were hatched — is correct. We have no record in historic times of their having spontaneously adopted a stream previously barren of them. Alewives retiu-n to the sea immediately after spawning (the old 'belief that they spawn but once and then die has no foundation), the spent fish on their return jom-ney to salt water being familiar sights in every alewife stream. After spaA\Ti- ing they are thin, but we have seen spent alewives that had already put on con- siderable fat taken from a trap at ProvincetowTi as early in the season as July 16 in 1915. Food. — The alewife, like the herring, is chiefly a plankton feeder, copepods, amphipods, shrimps, and appendicularians being the chief diet of specimens exam- ined b}' Vinal Edwards and Linton at Woods Hole. However, they also take small fish, such as herring, eels, launce, cunners, and their own species, as well as fish eggs. Unlike the herring, alewives often contain diatoms even when adult. Alewives fast when they are running upstream to sjjawn, but when the spent fish reach brackish water on their return they feed ravenously on the shrimp that abound in the tidal estuaries and which they can be seen pursuing. Commercial imporUince. — Alewives are excellent food fish, preferred by many to the sea herring, and a favorite bait for cod, haddock, and pollock. They are of considerable commei'cial importance among the minor fisheries.'' 42. Blueback {Pomolobus aestivalis MitchiU) Alewife; Glut herring; Summer herring; Blackbellt; Kyack Jordan and Evennann, 1896-1900, p. 426. Description. — Bluebacks and alewives are often confused; even experienced fishermen who recognize the existence of the two separate fish can not always tell them apart, so closely do they resemble one another in general appearance. There is one infalhble mark, however, that distinguishes the "blackbelly" not only from the alewife but from the liickory shad and herring as well, and that is that the lining of its belly is black, or at least black-spotted, instead of smoky gray as in its relatives. Apart from this we need state only that it is a slightly more slender fish than the 13 For a brief period during the war alewife scales commauded a high price for use in the maufacturo of artificial pearls. FISHES OF THE GUIiF OF MAINE 111 Fig. 46.— Blucback {Pomolobus ^stivalis) a, Adult. 6, Egg. c. Larva, newly hatched, 3.6 millimeters, d. Larva, 4 days old, 5.2 millimeters, c,' Fry, 30 mUlimeters 112 BULLETIN OF THE BUBEAU OF FISHERIES alewife (its body three and one-half times as long as deep), though differing so little in this respect that the two probably intergrade; that the fins are lower (here, again, the difference is so slight as to be hardly dependable); its eyes are smaller; and its back is rather darker blue; but, we repeat, to make certain which fish is in hand open it, glance at the belly Uning, and no doubt will i-emain. Size. — The blueback is of about the same size as the ale^^^fe. It grows to a length of about a foot and averages about a half pound in weight when mature. Color. — Dark blue above, the sides and belly silvery, with coppery reflections at least in some waters. General range. — This is a more southern fish than the alewife, occurring off the Atlantic coast of North America from the Bay of Fundy and Nova Scotia to Florida. It is more mmierous south than north of Cape Cod, and like the alewife spends the greater part of the year in salt water but runs up into brackish and fresh water to spa^vn. Occurrence in the Gulf of Maine. — Although Massachusetts fishermen have recognized the existence of two distinct species of alewives at least since 1816, it is difficiilt to arrive at a just idea of the abundance and migrations of this fish in the Gulf, because when "bluebacks" are reported they sometimes turn out to be alewives, while we have heard the latest run of alewives called bluebacks even in streams where the true blueback also occurs. It is certain, however, that schools of the latter are to be expected anj^where between Cape Sable and Cape Cod, for we have seen them freshly caught at Yarmouth on the west coast of Nova Scotia; they are not uncommon in the Bay of Fundy, Huntsman having had specimens from St. John Harbor and Shubenacadie River; they are definitely reported from the St. Croix River, from Eastport, Bucksport, Dennys River, Casco Bay, Small Point, Freeport, and sundry other localities along the coast of Maine, and are taken generally around the shores of Massachusetts Bay, including Cape Cod. Large numbers of bluebacks are sent to market, schools often being seined off the Maine coast outside the islands during summer and early autumn. These are mostly the 2-year-old fish, not yet sexually mature, judging from the fact that they are usually reported as very fat. Weirs also make la^ge catches of bluebacks from time to time, and we have seen thousands of them taken from a trap near Gloucester in June. How far offshore the bluebacks may wander is unknown. A few fish were reported under this name from Georges Bank during the investigation of 1913, but whether they were actually bluebacks or alewives is doubtful. It is sufficiently established that the blueback appears in our streams two weeks to a month later than the alewife, and that in streams frequented by both the later runs are bluebacks and the earher ones are alewives. In the Gulf of Maine, at least, it is apparently confined as a spawner to brackish ponds connected with the ocean and to the larger rivers,'* nor does it run far above tide water. Habits. — -The spawning habits of the blueback do not differ in any important particular from those of the alewife, except that it does not spawn until the water " Along the south shore ot Massaehasetts, tor instance about the head of Buzzards Bay, bluebacks, like alewives, ran up small streams. FISHES OF THE GULF OF MAINE 113 is much warmer — -70° to 75° instead of 55° to 60°.^^ No exact information as to the time of spawning in northern New Enghxnd rivers is available. About Woods Hole this takes place in early summer in small ponds with an outlet to salt water, most of the females being spent, a few females and many males ripe, and others of both sexes still unripe in July. The eggs are about 1 mm. in diameter, sink like those of the alewife, and are adhesive. Incubation occupies only about 50 hours at a temperature of 72°, the newly hatched larvae averaging 3.5 mm. in length but growing ^vithin 24 hours after hatching to 4 mm., with the greater part of the yolk sac already absorbed. Within a month they are 30 to 50 mm. long, and already show most of the diagnostic characters of the adult. Evidently the young soon find their way down to the sea, for bluebacks of 50 mm. have been seined in abun- dance in Rhode Island waters late in July. Notliing whatever is known of their rate of growth there. The spent fish, like alewives, return to sea shortly after spawning. Probably these are the bluebacks taken at Woods Hole and north of Cape Cod in September and October. The winter home of our bluebacks is ~yyii '*"*^, ITi. ^^^^7 Fig. 46.^AduIt shad {Aiasa sapidissima) unknown; probably, like their relative the sea herring, they move out from land and pass the cold season near the bottom. We need only note further that the blueback is as gregarious as the herring or alewife; that it is equally a plankton feeder, subsisting chiefly on copepods and pelagic shrimp, as well as on young launce and, no doubt, on other small fish fry. In commercial use no distinction is made on our coast between the blueback and the more abundant alewife — it is equally useful for bait and human food. 43. Shad (Alosa sapidissima Wilson) Jordan and Evermann, 1896-1900, p. 4'27. Description. — ^The shad resembles the alewife in the fact that its body is much deeper than thick and that its belly is sharp edged with bony serrations. In all respects it is a typical herring, with soft rayed dorsal and anal fins of moderate size, the former situated above the ventrals and weU forward of the middle of the bodv. " The early development and larval stages of the blueback are described by Euntz and Radclifle (1918, pp. 87-134) . 114 BULLETIN OF THE BUREAU OF FISHERIES It has a deeply-forked tail and large and very easily loosened scales. Unlike the sea herring, however, the shad has no vomerine teeth — adults, indeed, have no teeth at all, although young shad have small ones in the jaws which may persist imtil the fish is a foot or more long. It is easily recognized, being the deepest bodied of our herrings, a third as deep as long, and further marked among its relatives by the fact that the upper jaw is deeply notched at the tip with the end of the lower jaw fitting into the upper when the mouth is closed. The mouth, too, opens back farther than in the alewives, and the edge of the lower jaw is straight, not concave, as in the latter. The under jaw does not project noticeably beyond the upper, as in the alewives and especially in the hickory shad. Furthermore, the lining of the shad's belly is white — neither gray as in the herring and alewife nor black as in the blueback. Size. — The shad is the largest of herrings that regularlj' visit our Gulf, growing to a length of 2^ feet. Adult males rim in weight from IJ^ to 6 pounds; females from 3}^ to 8 pounds. Shad are occasionally reported up to 12 pounds, and the older writers mention them as heavy as 14 pounds, but none so large has been credibly reported in the Gulf of late 5'ears. Color. — Dark bluish or greenish above, white and silvery on sides and belly, Fig. 47 — Larva of the shad (Alosa sapidissima), 17 days old with a dusky spot close behind the rear edge of the gill cover, and frequently with several indistinct dusky spots in one or two longitudinal rows behind it. General range. — -Atlantic coast of North America from the Gulf of St. Lawrence to Florida, and represented by a close ally or variety in the Gulf of Mexico. It also has been successfully introduced on the Pacific coast of the United States. It runs up rivers into fresh water to spawn. Occurrence in the Gulf of Maine. — When the first settlers arrived in New England, they found seemingly inexhaustible multitudes of shad annually running up all the larger rivers and many of the smaller streams from Nova Scotia to Florida, with the tributaries of the Gulf of Maine hardlj^ less productive than the Hudson or Delaware; but as one stream after another was rendered impassable, or at least very difficult for the fish to ascend, by the construction of dams near the mouths, the local stock of shad has diminished until now the Gulf of Maine stock is but a shadow of its former abimdance, a fact more than one writer has taken a melancholy pleasure in lament- ing. Since it is the present status of the shad with which we are now concerned, the following table of the shrinking catch in the Merrimac will be a sufficient illus- tration of this depletion." " From Stevensen (1899). FISHES OF THE GULF OF MAINE 115 Year Number of shad caught, reported, or esti- mated Year =»= Number of shad caught, reported, or esti- mated 1789 830,000 540, 000 365, 000 60, 000 1.942 2,139 130 1888 None. 1889 18 1835 1S90-1S9C None. 1865 1893 2,020 1871 1873 faveraeel 1894 ... 2,750 1880 .- - 1895 - ---- -- 94 1896 7 The only Gulf of Maine rivers to which shad resort regularly at the present time are the Annapolis, '•Petit Codiac, Shubenacadie, and St. John, tributary to the Bay of Fundy, and the St. Croix,'' Pleasant, Harrington, Penobscot, and Kenne- bec Rivers '' in Maine. In the year 1896, 290,122 shad were reported as caught in the Kennebec system; 900 in the Pleasant; 300 in the Harrmgton River; only 114 in the Penobscot and 12 in the St. Croix; 100 in the Piscataqua; and 7 in the Merri- mac. Since then the stock has fallea even lower, for in 1919 the catch in Washing- ton County, Me., which includes the St. Croix, Pleasant, and Harrington Rivers, was only 400 pounds, say 100 fish, assuming them to average 4 pounds, with only 131 pounds (30 to 40 lish) taken in the Penobscot River, 3,121 pounds (700 to SOO fish) in Penobscot Bay, and 178,434 pounds (about 4.5,000 fish) from the Kennebec, its tributary estuaries and neighboring shore line (Sagadahoc and Lincoln Counties), that is, only about one-sixth as many as in 1896. In 1919 the total inshore and offshore catch for American fishermen in the Gulf of Maine north of Cape Cod was about 460,000 pounds (about 115,000 fish). No statistics are available for the few shad caught in the Baj^ of Fundy that year, but in 1916-17 the catch of shad in the Bay of Fundy was about 365,000 pounds, with about 9,000 pounds more along the west coast of Nova Scotia. The shad in salt water. — The life of the shad in salt water has long been con- sidered something of a mystery, but evidence gradually accumulates to the effect that its movements there are analogous to those of the herring, and that it does not perform the extensive north and south migrations with which it was formerly credited. Commencing with the spent shad on their return to the sea '' we find the New Brunswick fish (no doubt the Nova Scotian, also) making their way to the head of the Bay of Fundy on their return to the sea to fatten until they become the " fall shad " that are locally considered the choicest of fish. Large spent shad — presumably fish that have spawned in the Kennebec — are regularly caught in September and October about Mount Desert, where they have been the object of " The St. Croix formerly supported a large stock of shad. For 8 or 9 years prior to 1915 none came, but shad were again fairly plentiful in 1915 to 1916, according to investigations made by H. F. Taylor of the U. S. Bureau of Fisheries. '8 Shad have been entirely extirpated from the Saco, where they were formerly plentiful. " The following notes are based largely on reports by reliable fishermen and on our own observations, which we have gathered from catches during our several years' work on the New England coast. 116 BULLETIN OF THE BUREAU OF FISHERIES a consillerable frozen-tish industry of late,^° as well as near the Isles of Shoals and off York Beach in August, vrhile it has long been known that shad are present 40 to 50 miles at sea off the Maine coast throughout the autumn. A southward movement of these Bay of Fundy and Kennebec spawners is the most reasonable explanation for the 3"early presence off Cape Ann, from mid-October until into December,^' of large shad running from 13^ up to 10 pounds (averaging about 5 pounds), that is, fish that have spawned during the preceding summer. As a rule they are not abundant; sometimes, however, the pollock-netters make large hauls of them, as in the autumn of 1915, when 135,000 poimds of these large fish were caught near Gloucester. Sometime in December the large shad vanish, where they winter still being a matter for conjecture. Probably they sink and move out beyond the limits of extreme winter chilling, which may lead them to the central basin of the Gulf, a suggestion yet to be confirmed by actual captures of shad in winter but in line with the prevalent view that the shad of the middle and south Atlantic coasts of the United States move offshore to pass the cold season on the bottom. The young shad of the year, produced in southern rivers, are believed to winter near the mouths of their parent streams and this probably applies to the Gulf of Maine also. The mature shad with ripening sexual organs reappear off the western shores of the Gulf of Maine in April and May, when a few are picked up by haddock- netters between Cape Ann and Portland; most often about Boon Island and the Isles of Shoals." So few shad now frequent the Merrimac that it is probable these "spring shad" are bound north to the Kennebec River or Bay of Fimdy. Except for odd belated individuals the mature shad are all in the rivers or at least close to their mouths by the 10th or 15th of June, not to reappear in the sea until July or August (p. 116). Schools of small immature shad from a foot long and half a pound in weight up to 2 or 2J^ pounds, not yet of breeding age, that is, corresponding to the "fat" herring 2 or 3 years old, are reported every year at Provincetown for a short period in June, are sometimes taken in the weirs at Beverly and Manchester in Massa- chusetts Bay in June,-^ and are met with more or less commonly all summer off Cape Ann and thence eastward, which corroborates the general belief of local fishermen that they move north and east toward the Bay of Fundy as the summer advances just as the "fat" herring do (p. 101). However, instead of keeping off- shore these immature shad (which, like herring of corresponding age, are very fat) congregate in the bays of the Maine coast, even running up into brackish estuaries though never into fresh water. In Casco Bay, for example, where they have long been fished for, 64,490 pounds of shad (probably "fat" fish) were caught in 1896, though by 1919 the local catch had dwindled to only about 12,000 pounds (not over » About 250,000 pounds have been brought into the local freezers yearly from 1913 to 1915. " It has also been suggested that these fish are migrants from the south, visiting the rich plankton pastures of the Gulf for food, an interesting possibility that the evidence yet at hand can neither prove nor disprove. '" A series of shad from that region examined by Welsh at various dates (April 25 to May 17, 1913) averaged precisely 5 pounds, both sexes represented, and all with well-developed sexual organs. - " Numbers of shad about 14 inches in length were caught in the traps at Magnolia and Beverly from June 20 to July 6, 1921. FISHES OF THE GULF OF MAINE 117 5,000 fish, assuming an average of 2]/^ pounds). They have also been reported in abundance near Cape Ann, off the Isles of Shoals, near Boothbay Harbor, and at. Herring Cove near Eastjiort, Me. Summary. — The evidence at hand suggests that shad, like herring, spend the first winter near the mouth of the river in v/hich they are hatched; that for two or three years, as immatures, they roam our coast in summer, wintering some- where offshore; and that finally, as mature breeding fish, they come inshore in spring, run up rivers from April to June to breed, return promptly thereafter to the sea to pass the late summer and autumn, fattening near the surface not far from land, and, like herring, they winter offshore in deep water. Food. — The shad, like other herrings, is primarily a plankton feeder. We have found shad taken in the Gulf of Maine in summer packed full of copepods (chiefly Calanus), and the stomach contents of fish from the Nova Scotian coast of the Bay of Fundy examined by Willey (1923, p. 11) consisted chiefly of the copepod genera Acartia and Temora Avith other smaller ones, Mysid shrimps, and the larval stages of barnacles. Shad are also known to feed as greedily on the pelagic euphausiid shrimps as herring do, on fish eggs, and even on bottom-dwelling amphipods, showing that at times they forage near the ground. They are not known to eat fish. Breeding habits.^* — It is now sufficiently established that on their spawning migration shad return year after year to the same general region, sound, or estuary; and in the Gulf of Maine, where so few rivers can now serve as spawning grounds, this necessarily means to the same stream, the date when the sexually mature shad enter fresh water being governed by the temperature of the streams — that is, when the river water has warmed to 50° to 55°. Consequently the shad "run" cor- respondingly later in the year passing from south to north along the coast. Thus the run commences in Georgia in January; in March in the waters tributary to Pamlico and Albemarle Sounds; in April in the Potomac; and in May and June in northern streams generally from the Delaware to Canada. In the Kennebec, according to Atkins (1887), the first sliad appear late in April, with the main run in May and June; the first ripe females are caught the last week in May and they begin to spawn about June 1, most of them doing so during that month, a few in July, and possibly an occasional fish as late as August. Probably these dates applied equally to the Merrimac in the good old days when shad were plentiful there, but the season begins somewhat later in the St. John, as might be expected, with the fish running from mid-May until the end of June. The fish select sandy or pebbly shallows for spawning ground. On the average, females produce about 30,000 eggs, though in the case of very large fish as many as 156,000 have been estimated. After spawning the spent and very emaciated fish at once begin their return journey to the sea. In the Kennebec they are first seen on their way down about June 20 and constantly thereafter throughout July; in the St. John spent fish are running down in July and August. According to " Accounts of the breeding habits of the shad have been given by Ryder, Report U. S. Commissioner of Fish and Fisheries, 1885 (1887); by Prince (1907); and in the Manual of Fish Culture. 1900. 118 BULLETIN OF THE BXJEEAU OF FISHERIES Atkins they begin feeding before reaching salt water and recover a good deal of fat before moving out to sea. The eggs are transparent, pale pink or amber, and being semibuoyant and not sticky like those of other river herrings they roU about on the bottom with the current. The period of incubation is from 5 to 10 days in the temperatures prevailing in June in the Kennebec and St. John Rivers, having been found by experiment to be as follows: Water temperature Period of incubation 74° F. (23.3° C.) 70 hours, or about 3 days. 64° F. (17.8° C.) . 109 hours, or about 4H days. 57° F. (13.9° C.) 148 hours, or about 6 days. 54° F. (12.2° C.) 408 hours, or about 17 days. The larvfe are about 9.5 mm. long at the time of hatching, growing to 15.5 mm. by the ninth day with disappearance of the yolk sac. At 21 to 28 days the fins are fully developed and the fry have attained a length of about 20 mm. Shad larvae much resemble herring, being extremely slender with the vent almost as far back as the base of the tail. The young shad remain in the rivers until fall when, at a length of 1}4 to 43^ inches and resembling their parents in appearance, they move down to salt water. The length attained by the shad during its first autumn depends on the date of hatching. In the rivers of Maine the fry may be as long as 5 to 7 inches by the first week in November; even larger (6 to 8 inches) in New York streams. Nothing definite is known of the rate of growth of the shad after it leaves its parent river; presumably, however, it grows little during the first winter but is about a foot long by the second autumn. It is supposed to mature at about the same age as the alewife^S or 4 years 44. Menhaden (Brevoortia tyrannus Latrobe) POGY Jordan and Evermann, 1896-1900, p. 433. DescHption, — This fish is universally called "pogy" in the Gulf of Maine but no less than 30 common names are in use south of Cape Cod. Like all our other herrings it is flattened laterally, has a sharp-edged belly, and is as deep propor- tionally as the shad (body three times as deep as long), though when the fish are fat the general form is altered. The very large scaleless head, which occupies nearly one-third of the total length of the body, gives the menhaden an appear- ance so distinctive that it is not apt to be mistaken for any other Gulf of Maine fish. It is likewise distinguishable from all its local relatives by the fact that the rear margins of the scales are vertical — not rounded as in the more typical herrings — and edged with long comblike teeth instead of being smooth. The dorsal fin, furthermore, originates slightly behind the ventrals and is thus posterior to the latter for the whole length, whereas in herring, alewives, and shad it stands directly over the ventrals. We need only point out further that the pogy is toothless, its FISHES OF THE GULF OF MAINE 119 tail deeply forked, its ventral fins very small, its dorsal and anal of moderate size, its mouth large and gaping back as far as the hind margin of the eye, and that the tip of its lower jaw projects beyond the upper. Fig. 48. — Menbaden (Brcioortia tyrannus) a, Adult, b. Egg. c, Larva, newly hatched, 4.5 millimeters, d, Fry, 23 millimeters, c, Young flsh, 33 millimeters. Size. — Adult menhaden average 12 to 15 inches in length, and from two-thirds to 1 pound in weight. One 18 inches long was taken at Woods Hole in 1876, and a fish 20 inches long has been reported. 120 BULLETIN OF THE BUREAU OF FISHERIES Color. — Dark blue, blue gray, or blue brown above, with silvery sides, belly, and fins, and a strong yellow or brassy luster. There is a conspicuous dusky spot on each side close back of the gill opening; with a varj^ing number of smaller dark spots behind it. General range. — Atlantic coast of North America from Nova Scotia to Brazil. Occurrence in the Gulf of Maine. — The Gulf of Maine is the northerly limit of range of the menhaden; St. Mary Bay on the west coast of Nova Scotia is its most easterly outpost. Prior to about 1850 the pogy seems not to have been un- common at the mouth of the Bay of Fundy; it was, indeed, reported by Perley (1852) as far up the bay as St. Jolm, and fishermen spoke of it as abundant near Eastport up to 1845 or 1850. Since then, however, it seems to have abandoned Fundian waters altogether " except for an occasional straggler, and very few menhaden have been noticed east of Mount Desert and Jonesport of late years. Perhaps the most interesting aspect of the occurrence of the menhaden in the Gulf of Maine is that it fluctuates tremendously in abundance from year to year, periods of great plenty alternating with periods of scarcity or entire absence from our waters. Thus 1845 was a "big year," while in 1847 pogies were very scarce. Then for some years prior to 1875 they were tremendously abundant off the coasts of Massachusetts and Maine every summer, and a considerable men-, haden fishery grew up on the Maine coast. Since then the local stock has under- gone the most violent fluctuations imaginable, of which abundant testimony is to be found in the files of the Bureau of Fisheries. Thus very few menhaden were taken in the Gulf during the cold summer of 1877 until September and October, when they were reported as about as abundant as normal. Practically none appeared north of Cape Cod in the year 1879, as striking an abandonment of a considerable area by a fish previously abundant there, perhaps, as has taken place within recent times. For the next six years menhaden were so scarce along the coast of Maine that when odd ones were picked up in the weirs or were seined it caused cornment (in 1883, for instance, a few were reported to the bureau but no schools were seen), and many people thought they had gone for good; but in 1886 they were once more reported abundant off Maine and Massachusetts, while in 1888 they were so plenti- ful as far east as Frenchmans Bay that the menhaden fisheries were revived. Menhaden were as plentiful in Maine waters in 1889 as they had ever been, with more than 10,000,000 pounds taken there, and they were still so numerous in 1890 that four fertilizer factories were established and nearlj^ 90,000,000 fish were taken during the season. This period of abundance was short-lived, however, less than half as many fish being caught in Maine waters (about 41,000,000) in 1891 as in 1890, while few menhaden were taken or seen north of Cape Cod in 1892. In 1894, however, the fish were once more sufficiently abundant in the Gulf of Maine for a single steamer to seine about 1,000,000 fish off the Kennebec during July and the first weeks of August, and 582,131 fish were taken in Boston Harbor in 10 days' fishing during the last half of that month. " According to Huntsman (1922a) one was taken in St. John Uarbor in .\ugust, 1919. FISHES OF THE GULF OF MAINE 121 During the period 1895 to 1897 menhaden were scarce in the Gulf, but abundant again in 1898, about 7,000,000 pounds being taken along the Maine coast. In 1902 the Maine catch fell to only about 300,000 pounds. In 1903, however, they were again reported as very abundant north of Cape Cod, especially in Boston Harbor; but from 1904 to 1921 menhaden were rare north of Cape Cod, and some years a few schools were seined in Massachusetts and on the Maine coast while during other summers very few were seen. After 20 years of scarcity, however, they reappeared in great abundance in the southwest part of the Gulf in the sunmier of 1922, and by the first week in August many schools had been seen along the outer shore of Cape Cod. Eighteen steamers fished successfully for some weeks in Massa- chusetts Bay, and menhaden were reported as plentiful at least as far north as Boothbay Harbor, where about 2,500 barrels were frozen. The fact that upwards of 1,500,000 pounds were landed by the larger fishing vessels besides what the small boats brought in will give some idea of their abundance in the 'Gulf during the summer in question, but no large schools were reported east of Boothbay Harbor. The appearance of menhaden in such abundance in the Gulf after so many years' absence prompted the Bureau of Fisheries to send the steamer Halcyon to Massachusetts Bay in August for an investigation of this phenomenon. ,V preliminary examination of the towings revealed the presence of much greater quantities of diatoms than is usual at that season, showing that the fish found a better pasture in Massachusetts Bay than in any summer since 1912, but this evidence hardly warrants the definite conclusion that it was an unusually rich food supply that attracted them past Cape Cod. The menhaden is a summer fish with us, all reports agreeing that in its years of plenty it appears about mid-May in Massachusetts Bay and during the last weeks of May or the first part of June off the Maine coast, and that more and more continue to appear for a month after the first fish arrive. Fogies are most abundant during July, August, and in early September, after which few are seen. Most of them depart from the coast of Maine by the middle of October, and from the Massa- chusetts Bay region by early November, while by the middle of that month it would be unusual to find a single menhaden along these shores. The universal belief among fishermen that the seasonal appearance and dis- appearance of menhaden in the Gulf of Maine result from a definite migration from the south around Cape Cod in the spring and a return journey in the autumn is probably well founded, for, unlike the herring, it is a warm-water fish, and our study of the temperature of the Gulf of Maine corroborates earlier observations to the effect that it never appears in spring until the coastwise water has warmed to 50° or more, or in abundance until the temperature is several degrees higher than this, which is in accord with Bean's (1903) experience that menhaden will not survive in an aquarium if the water chills below 50°. No doubt it is the falling temperatm-e of autumn that forces the menhaden to leave the bays of northern New England. It is generally believed that the Gulf of Maine fish round Cape Cod and travel westward in their autimin migrations-' as far as the eastern end of Long Island, 2« Smith, 1896, p. 299. 122 BULLETIN OF THE BUREAU OF FISHERIES wliere they have been thought to leave the coast and proceed directly out to sea, but this is not established nor is their winter home known. The menhaden usually hug the shore in summer, most of the catch being taken inside our bays, in the outer harbors, or at the farthest not over half a dozen miles from land, and though this rule has its exceptions — some j'ears they congregate chiefly as much as 40 to 50 miles oiTshore, 1878 being an instace in point — we have heard no report of menhaden in the central part of the Gidf or on Georges Bank. In menhaden years the fish occur all along the shores of the Gulf of Maine from Cape Cod to Penobscot Bay, even to Mount Desert, but their chief centers of abundance always lie in Massachusetts Bay within a mile or so of land, par- ticularly off Barnstable and in the mouths of Boston and Salem Harbors, in Casco Bay, among the islands, and thence to Penobscot Bay. Breeding habits.-'' — Menhaden spa\vn all along the eastern coast of the United States as far north as the southern part of the Gulf of Maine, the breeding season varying with latitude. Thus spa^vning occurs in late fall and early winter on the south Atlantic and Gulf coasts, but oft' the middle Atlantic States menhaden spawn in summer and through the autumn, while captures of eggs and of larviB about Woods Hole prove that spawning takes place there chiefly in June and continues until well into October. The menhaden is equally a summer spawner in the Gulf of Maine, where spent fish and others approaching maturity have been reported during July and August. Up to the present, however, we have found no eggs in our tow-nettings north of Cape Cod (though yoimg fry were taken in abundance in Casco Bay in October, 1900), probably because our work has been carried on during a series of poor menhaden seasons. Menliaden eggs are buoyant and resemble those of the European pAcliavd (Clupea pilchardiis), but are easily distinguished from the eggs of any other Gulf of Maine fish by their large size (1.5 to 1.8 mm. in diameter), broad perivitelline space, small oil globule (0.15 to 0.17 mm.), and very long embryo. Incubation, as Welsh found by experiment, is very rapid (less than 48 hours) . The newly hatched larvae are 4.5 mm. in length, growing to 5.7 mm. in four days after hatching. The dorsal and caudal fins first become visible at a length of 9 mm. ; at 23 mm. all the fins are well developed; at 33 mm. scales are present; and at 41 mm. the fry show most of the characters of the adult though their eyes are proportionately much larger. The youngest larvae much resemble young herring, but the fins are formed, the tail becomes forked, and the body deepens at a much smaller size, a menhaden of 20 mm. being as far advanced in development as a herring of 35 mm., which makes it easy to distinguish the older larvae of the two fish. Welsh concluded from examination of great numbers of fry and from measure- ments and scale studies of fish of various ages that menhaden spawned in summer (which would apply to most of the fry produced in the Gulf of Maine) are 6 to 8 cm. (234 to 3}4 inches) long their first winter and average slightly more than 16 cm. (about 6J^ inches) the second winter, while faU-spawned fish are 3 cm. (134 inches) and about 13 cm. (about 5 inches) long in their first and second winters, respectively, -with every gradation between the two, depending on the precise season when the "' Tbe breeding habits ot the menhaden are described by Kuntz and EadoliiTe (1918, p. 1191 . FISHES OF THE GULF OF MAINE 123 fish are spawned. Sexual maturity is apparently attained in the season following the third winter, and a few of the older fish that he examined showed as many as^9 to 10 winter rings on the scales. Wlien the menhaden first arrive on our coasts in spring they are thin, but they put on fat so rapidly that while the average yield of oil per 1,000 Gulf of Maine fish was about 12 gallons for the whole summer season of 1894, it rose to 143/^ gallons for Boston Harbor fish in August and to 16 or 18 gallons in September. It is generally accepted, furthermore, that fish taken on the New England coast always average larger and fatter than those caught farther south. Food. — The menhaden, formerly thought to subsist on mud, is now known to feed chiefly on microscopic plants, particularly diatoms, and on the smallest Crus- tacea.^* These it sifts out of the water with a straining apparatus in the shape of successive layers of pectinated gill rakers as efficient as our finest nets. Men- haden feed, as Peck described, by swimming with the mouth open and the gill openings spread, and we have often seen specimens in the aquarium at Woods Hole doing this.^° The mouth and pharyngeal sieve act exactly as a tow net, retaining whatever is large enough to enmesh with no voluntary selection of par- ticular plankton units. The prey thus captured, as appears from the stomach contents, includes small annelids, various minute Crustacea, schizopod and decapod larvsB, rotifers, etc., but as a rule these are greatly outnumbered by the sundry unicellular plants, particularly by diatoms and peridinians. At a given locality the food eaten parallels the general plankton content of the water, except that none of the larger animals, on the one hand, nor the verj' smallest organisms (that is, certain infusoria), on the other, appear in the stomachs of the fish. The men- haden, in short, parallels the whalebone whales in its mode of feeding, except that its diet is finer because its filter is closer meshed. Peck has calculated from obser- vations on the living fish that an adult menhaden is capable of filtering between 6 and 7 gallons (about 24 to 28 liters) of water per minute, and while the fish do not feed continuously this will give some measure of the tremendous amount of water sifted and of plankton required to maintain the hordes in wliich these fish appear. The abundance of microscopic plants in the water of bays, estuaries, etc., has often been invoked to explain the concentration of menhaden close to the shore. Enemies. — No wonder the menhaden, fat and oily, swimming as it does in great schools of closely ranked individuals and helpless to protect itself, is the prey of every predaceous animal that swims, and that the havoc wreaked on it by other fish has often been described. Whales and porpoises devour them in large numbers; sharlcs are usually seen following the pogy schools; pollock, cod, silver hake, and swordfish all take their toll in the Gulf of Maine, as do weakfish and bluefish south of Cape Cod. Tuna, or "horse mackerel," kill great numbers, but the worst enemy of all is the bluefish, and this is true even in the Gulf of Maine during periods when both bluefish and menhaden are plentiful there (p. 239). Not only do these pirates devour millions of menhaden every summer but they kill far " For a detailed account of the food and of the branchial sieve of the menhaden, see Peck (Bxilletin, United States Fish Com- mission, Vol. XIII, 1S93 (1S94), pp. 113-124, pis. 1-8. Washington). " Apparently Ehrenbaura (as quoted by BuUen, Journal, Marine Biological Association of the United Kingdom, Vol. IX, 1910-13, pp. 394-403. Plymouth) was not acquainted with the habits of the menhaden when he wrote to the effect that no fish eat plankton indiscriminately or habitually swim about with open mouth when feeding. 124 BULLETIN OF THE BUKEAU OF FISHERIES more than they eat. Besides the toll taken by these natural enemies, menhaden, like herring, often strand in myriads in shoal water either in their attempt to escape their enemies or for other reasons, to perish and pollute the air for weeks with the stench of their decaying carcasses. Habits. — The menhaden, like the herring, almost invariably travels in schools of thousands of individuals, swimming closely side by side and tier above tier when, as Goode, et al. (1884, p. 571) so graphically write, "one may see their glittering backs beneath, and the boat seems to be gliding over a floor inlaid with blocks of silver." In calm weather menhaden come to the surface, where fishermen recog- nize the identity of the schools by the ripple they make. W. F. Clapp has described the visible difference between menhaden, herring, and mackerel, as follows: Pogie.'!, like herring, make a mvich more compact disturbance than mackerel, which are often much scattered. Fogies make a much bluer and heavier commotion than herring, which hardly make more of a ripple than does a light breeze passing over the water. Besides, the indi- vidual pogies or herring seldom show themselves, whereas mackerel often break the surface with their heads while swimming. It is chiefly on warm, still, sunny days that the menhaden come to the sur- face — sinking in bad weather — and they are said to come up more often on the flood tide than on the ebb. It is also said — but this we can not vouch for — that the fish work inshore on the flood tide and offshore on the ebb. Commercial imporfance. — Conmiercially the menhaden is one of the most impor- tant of our American fishes— not for the table, but for the manufacture of oil and fertilizer'"' — but, as pointed out above (p. 120), it is only in certain years that a large catch is made north of Cape Cod. The fact that the total value of menhaden products in the year 1912 was $3,690,155 will give an idea of the magnitude of the industry. Practically the entire catch of menhaden is taken with purse seines; they never bite a baited hook. Menhaden are used to a very limited extent for food, but so oily a fish is never likely to become popular. THE ANCHOVIES. FAMILY ENGRAULIDID.^ The anchovies are small herringlike fishes, readily distinguishable from the latter by the fact that the mouth is not only very much larger and gapes much farther back, but is inferior in situation rather than terminal, and is overhung by the upper jaw, which projects like a short piglike snout in some species. Only one anchovy, a straggler from the south, is known to occur in the Gulf of Maine. 45. Anchovy {Anclwvia mitchilli Cuvier and Valenciennes) Whitebait Jordan and Evermann, 1896-1900, p. 446. Description. — The only Gulf of Maine fishes with which one might confuse the anchovy are young herring, smelt, or silversides, but it may easily be distinguished from the former by the wide mouth, as just noted, by the fact that the upper jaw overhangs the lower instead of vice versa, by its much larger eye, by the relative " For an account of the status of the menhaden industry in 1912, see Qreer (Appendix lU, Report, U. S. Commissioner of Fisheries, 1914 (1915), 27 pp.). FISHES OF THE GULF OF MAINE 125 positions of the fins, the dorsal being altogether behind instead of over the ventrals, •nath the latter originating close behind the tips of the pectorals when these are laid back against the body, by its much longer anal fin, and by the fact that the belly is rounded instead of sharp edged. The lack of an adipose fin back of the dorsal is sufHcient to separate anchovy from smelt at a glance, while the silversides (Menidia) has two dorsal fins instead of only one. The anchovy has large, thin, easily detached scales and a deeply forked tail. Its body is about four times as long as deep, and compressed, while the tip of the upper jaw or "snout" is shorter than in most of its immediate relative^. Fig. 40,— Egg • / ■„ Fig. 50.— Larva, 10 millimeters m»- Fig. 61.— Adult ANCHOVy (Anchovia mitchilU) Color.— The anchovy is a whitish silvery translucent little fish, its most charac- teristic marking being an ill-defined silvery band scarcely wider than the pupil of the eye runmng from the gill opening back to the caudal fin. There are also many dark dots on body and fins. Size. — Two and one-half to four inches long. General range.— Coast of the United States from Maine to Texas, chiefly west and south of Cape Cod. Occurrence in the Gulf of Maine.— The anchovy is mentioned here because it has been taken in Casco Bay and at Provincetown. It has no real place in the Gulf of Maine fauna, being a southern fish that rarely strays past Cape Cod, though it is abundant about Woods Hole and thence westward and southward. Stragglers 126 BULLETIN OF THE BUREAU OF FISHERIES may be expected most often in the Gulf in midsummer since it appears from May to October in southern New England waters. Sandy beaches and the mouths of rivers are its chief resorts. Habits. — An accoimt of its embryology and larval development is given by Kuntz." THE SALMONS. FAMILY SALMONID.E The salmons are soft^rayed fishes with no spines in any of the fins, with the ventrals situated on the abdomen far behind the pectorals, and with a fleshy rayless "adipose" fin on the back behind the rayed dorsal fin, the presence of this adipose fin and its location separating them from all other Gulf of Maine fishes except the smelt family, the pearlsides (p. 151), and the viper and lancet fishes (p. 155).^^ The rounded noses, stout bodies, and nearly square tails of the salmons mark them at a glance from the sharp-nosed, slender, forked-tailed smelts; the absence of phos- phorescent organs distinguishes them from the pearlsides, while the viper and lancet fishes are of quite different general aspect. At the present time three salmons ^^ occur in the Gulf of Maine, one of which — the sea trout — resorts to tidal estuaries at the mouths of a few of our streams, while a second — the humpback salmon — has recently been introduced from the Pacific coast (the success of the experiment is still in doubt), leaving the Atlantic salmon alone as a characteristic inhabitant of the open waters of the Gulf of Maine. KEY TO GULF OF MAINE SALMONS 1. Anal fin long, with 14 to 17 rays Humpback salmon, p. 126 Anal fin short, with less than 13 rays 2 2. Scales so small as to be hardly visible; back with vermiculate markings; teeth on roof of mouth confined to a group in front Sea trout, p. 138 Scales large enough to be easily visible; back without vermiculate markings; a row of teeth runs back along the midline of the roof of the mouth Salmon, p. 130 46. Humpback salmon iOncorhynchus gorhuscha Walbaum) Jordan and Evermann, 1896-1900, p. 47S. Description. — The humpback is of the familiar salmon outline while living in the sea, the body being deeper than thick but with rounded belly. (See fig. 52, p. 129.) The head is naked but the body is covered with scales large enough to be seen easily. Its dorsal fin stands about midway of the body above the ventrals, and the flaplike adipose fin is over the rear end of the anal. In all this it agrees so closely with the Atlantic salmon that the two might easily be confused were it not that the anal fin of the humpback invariably has 14 or more rays while that of the Atlantic sahnon has only about 9 . The male humpback, Uke all the Pacific salmons and to a lesser degree the Atlantic salmon, undergoes a very noticeable change in form in the spawning season , when the body deepens and develops a prominent hump in front of the dorsal fin, while the jaws elongate and become hooked at the tip and the teeth increase in size. " Bulletin, U. S. Bureau of Fisheries, Vol. XXXIII. 1913 (1915), p. 13. " Sundry other deep-sea fishes have adipose fins. " A specimen of one of the whiteflshes (probably Coregonun quadrilateral^ Richardson) was taken in the mouth of the Sissibou River, St. Mary Bay, September, 1919 (Uuntsman, 1922a, p. U), straying down from fresh water. Whiteflsh are recognizable by the presence of an adipose fin, a."! in the true salmons, but a very small mouth and compressed, herringlike rather than salmon- like body. FISHES OF THE GULF OF MAINE 127 Color. — While in the sea the back and tail of the humpback are bottle green with poorly defined black spots. These spots are particularly conspicuous on the tail, where they are oval in outline and as much as a third of an inch across the long diameter. These large oval spots on the tail form one of the most distinctive marks whereby the humpback can be distinguished from all other salmon. Its sides and belly are silvery, with a faint pinkish tinge. Young humpbacks are unique among salmon in being of practically adult coloration without "smolt" marks (p. 133). Size. — The humpback is the smallest of the Pacific salmons and is much smaller than the Atlantic salmon, adults averaging only about b}4, pounds in weight and 20 to 25 inches in length, males running up to about 11 and females to 73^ pounds. General range. — Pacific coast and rivers of North America and Asia, from Oregon northward on the American side. This is the most abundant salmon in Alaska. It runs up fresh rivers to spawn, which it does but once and then dies. Occurrence in the Gulf of Maine.— The history of the introduction of this west coast salmon to New England waters is as follows: In the autumn of 1913 a large consignment of humpback eggs was shipped to the Craig Brook and Green Lake (Me.) hatcheries, and the approximately 7,000,000 fry and fingerlings hatched therefrom were distributed in the Penobscot, Andro- scoggin, Damariscotta, Dennys, Pleasant, Union, Medomak, Georges, and St. Croix Rivers. A year later some 5,000,000 more young fish were liberated. A third plant was made in 1915, a fourth of 6,235,808 fingerlings in 1916, and a fifth of about 1,000,000 in the Dennys and Pembroke Rivers in 1917.^* The results of this attempt at acclimatization were first seen in the summer and fall of 1915 when fishermen along the Maine coast reported large numbers of mature humpbacks. Furthermore, humpbacks ran in the Dennys River (where many were caught) from August 15 until September 24, and some probably spawned there, for the bodies of spent fish were seen drifting downstream. Himipbacks again entered the rivers of eastern Maine, particularly the Pembroke and Dennys, with a few reported from the Penobscot, St. Georges, Medomak, and St. Croix, during August, September, and October, 1917, the result of the plant of 1915. In the Dennys alone at least 2,000 matm-e fish were seen and many averaging about 5 pounds and one as heavy as 10 pounds 9 ounces were caught. Definite information for 1918 is lacking, but even larger numbers entered the Dennys and Pembroke Rivers in the autumn of 1919 than in 1917, with smaller runs in the Penobscot, Machias, St. Croix, and Medomak Rivers, and humpbacks were caught in weirs in Passamaquoddy and Cobscook Bays near the mouth of the Bay of Fundy. Enough spawned in the Dennys and Pembroke Rivers that year for the fish-culturists of the Bureau of Fisheries to artifically fertilize half a million eggs. In 1920,'^ too, adult fish wei-e taken in the weirs in Penobscot Bay, and some time during the summer of 1921 one fish was caught in a weir as far from its native river as LanesviUe, Mass. (near Cape Ann), whence it was forwarded to the Massa- chusetts commissioners as reported by C. E. Grant, of Gloucester. " More detailed accounts of the successive plantings will bo found in the annual reports of the Commissioner of Fisheries for the years 1914 to 1920. " Reported catch, Washington County, Me., 1920, 310 pounds. 102274— 25 1 9 128 BULLETIN OF THE BTJEEATJ OF FISHERIES The experiment has gone far enough to prove that the humpback can live and grow in the Gulf of Maine during its normal period of marine existence and find its way back to the home streams as the time of sexual maturity draws near. Whether it can multiply to any considerable stock, or even maintain itself by na- tural reproduction -without the aid of artificial propagation in the few Maine rivers open to it, remains to be seen. Brief experience Avith it in the Gulf of Maine sug- gests that the local run vnll take place in late summer and early autimin as is the case in Alaska. Habits. — The humpback runs up small streams indifferently, whether or not lakes occur in their courses. After it enters fresh water it feeds no more, and its digestive organs, like those of all Pacific salmons, shrivel up and atrophy while the changes in form of body and jaws, so characteristic of the breeding male (p. 126), take place. The skin, too, thickens and becomes so spongy that the scales are entirely concealed. The humpback spawns very soon after it enters the rivers, and though it reaches the spawning grounds in fairly good condition (except for a loss of fat), the fins of both sexes soon become frayed, the skin rubs off the jaws, bases of fins, and other prominent places, the tails of the females are worn down to the quick, fungus attacks these open wounds as well as the gills and eyes, and when the last eggs and milt are deposited the spent and exhausted fish finally die. No hmnp- back ever survives the operation of spawning. Spawning takes place in the faU at temperatures of 54° to 60° F. As soon as the yolk sac of the young humpback is absorbed and it is able to swim (which is about the time the ice breaks up in spring) it runs dovm to the sea. During their first months in salt water the fry linger near the mouths of the home streams, ^° where they feed chiefly on copepods and other small crustaceans, on pteropods, and on insects that drift downstream with the current, and occasionally on fish fry. After they are 5 or 6 inches long they move out into deep water, and very little is known of their habits and wanderings thereafter until they reappear on the coast as adults to breed. Large humpbacks have been found full of pelagic Crustacea and launce, evidence that they subsist on a mixed plankton and fish diet, the former probably predominating. No humpback has ever been known to take a trolled spoon or baited hook. McMurrich " (in 1912) and Gilbert'* (in 1913) have proved by their studies of its scales that the humpback invariably lives in the sea through one summer, a winter, and well into the secOnd summer, and then comes in to spawn. Commercial importance. — The humpback is an excellent food fish when taken in salt water and would be a valuable addition to the Gulf of Maine, but it becomes worthless soon after entering fresh water.'' >« We owe to Chamberlain (Bureau of Fisheries Document No. 627, Report, U. S. Commissioner of Fisheries, 1906, 112 pp.. Pis. I-V) the little we know of the habits of the young humpback. " Proceedings and Transactions, Royal Society of Canada, Third Series, Vol. VI, May, 1912 (1913), Section IV, pp. 9-28. M Bulletin, U. S. Bureau of Fisheries, Vol. XXXII, 1912 (1914), pp. 3-22, Pis. I-XVII. Washington. 39 The Pacific "Chinook" salmon (Oncorhynchus tschawytscha) was introduced in the Merrimac River system by the Com- missioners of Fisheries and Game of the State of Massachusetts in 1916, but the plant seems to have been a failure, for no adults w 1916-17 ' Nova Scotian shores of the Qulf of Maine Pounds 74, 523 156, 5TS 43, 500 Pounds 264, 900 381, 500 83, 400 Pounds 152, 740 139 Pounds ' 20, 000 New Bnmswick shore of the Bay of Fundy. New Brunswick rivers tributary to the ■Ray nf FnnHy Coast of Massachusetts north of Cape Cod MOO 1 The Canadian and United States returns for 1889 are directly comparable, both being for the calendar year, but the Canadian returns for 1916-17 were for the fiscal year Mar. 31 to Mar. 31. ' The Canadian catch was larger the year previous. 3 Approximate. A catch of 250,000 to 300,000 pounds (about 25,000 to 30,000 fish) may now be expected annually along the Canadian shores and in the Canadian rivers of the Gulf; in New England waters less than 20,000 pounds. Fishery by modern methods no doubt would have yielded a very much larger total in past times when the fish ran in all the New England rivers. Although no salmon now spawn south of the Penobscot a few are still taken every S limm er in the weirs in the Massachusetts Bay region. Up to about 1895 Cape Cod Bay annually yielded a number of large fish weighing up to 25 pounds or more, as well as many young smolts of about 6 inches (p. 133), with 1892 as a particularly productive year. A few fish were taken yearly about Cape Ann, also, and in Cape Cod Bay until about 1908. Thereafter the Massachusetts catch fell practically to nil. In some seasons an odd fish or two was reported from one place or another in the Massachusetts Bay region, other years none at all," imtil 1917 when perhaps 150 small salmon of from 2 to 5 pounds were caught by the mackerel gill-netters in Ipswich Bay and off Thachers Island during November and early December. " Radcliffe, 1921, p. 146. " For further data on salmon in Massachusetts waters see the annual reports of the Massachusetts commissioners since 1902. FISHES OF THE GULF OF MAINE 133 Life history. — -It is no wonder that the life of as noble a fish as the salmon has been the subject of much scientific study and that a whole literature has grown up about it. As everybody knows the salmon lives the greater part of its life and makes most of its growth in the sea but spawns in fresh water. In Gulf of Maine rivers this happens in October and early November on sandy or gravelly bottom far upstream, the females smoothing a shallow trough and covering the eggs with gravel. The large (6 to 7 mm.) thick-shelled eggs lie loose on the bottom and develop so slowly in the low temperature of winter that hatching does not take place until late in the following April or early in May. The newly hatched larvae are 15 to 18 mm. (0.6 to 0.7 inch) long, and carry a tremendous large yolk sac for about six [weeks, hiding among the pebbles of the spawning bed and taking no food. When the yolk sac is absorbed the little fish begin to swim and feed. These little salmon, known as "parr," soon assume a brilliant coat with 10 or 11 dark trans- verse bars alternating with bright red spots, much like a young trout. Parr live in fresh water for longer or shorter periods, according to locality or to other factors not well imderstood. In the rivers tributary to the Gulf of Maine probably most of them remain two winters and two summers, running down to sea the third summer, but some may seek salt water toward the end of the second summer (when 1)^ years old), as is the rule in rivers tributary to the Gulf of St. Lawrence. Parr may be moving downstream any time from late spring to autimm, but in Gulf of Maine streams most of them probably make the journey in June and July, being then 5 to 6 inches long. On the other hand it is possible that some may linger in fresh water until 3, 4, 5, or even 6 years of age, as is known to happen in Norway. An occasional male parr may even become sexually mature before visit- ing the sea at all. As they near tidewater the parr put off their barred and spotted pattern to assmne the silvery coat worn by the salmon during Ms sojourn in the sea. They are then known as "smolts." After they reach salt water they are found for a time about the shores of estuaries and in river mouths. No doubt the little salmon, too small to sell, that are caught in summer and autumn in weirs at Matinicus Island come from the Penobscot a month or two previous, but they leave the coast in autumn and little is known of their movements during the first winter. Smolts from 7 to 12 inches in length have occasionally been taken in drift nets in Nor- wegian waters and in the North Sea with herring, mackerel, and other fish, sug- gesting that they follow and prey upon the shoals of the latter, but there is no reason to suppose that they move far offshore. However this may be, they grow so rapidly on the abxmdant diet the sea affords that they usually reach a length of at least 16 inches and weigh anywhere from IJ-^ to 7 or more pounds by the fol- lowing spring when they reappear on the coast. They are now termed "grilse," distinguishable from the older fish by a more forked tail, more slender form, smaller head, thinner scales, and rounded and more numerous spots that are bluish rather than black. Some of the male "grilse" become sexually mature, and although fewer female salmon mature until older, "gi-Use" of tliis sex (fish only one year at sea) accompany the males and the older mature fish up the rivers. ' In northern Canadian streams gi'ilse are very abundant. This is also the case in the St. Johns 134 BULLETIN OF THE BUREAU OF FISHERIES but for some unknown reason it appears that so few fish return to fresh water along the coast of Maine untQ older that in past times when there was still a good run of salmon not more than 3 or 4 grilse to 70 adults were taken in the St. Croix, and not more than 1 to 500 in the Dennys and Penobscot Rivers. Salmon, like most other northern sea fish, make most of their growth diu-ing the siunmcr. In summers when they spawn they hardly grow at all, but unlike the herrings, cod, etc., all mature salmon do not spawn. Hence the size of a salmon depends more on the number of times it has spawned than on its age. Some males, as just remarked, spawn after one year at sea; most of them, and many females, after two; while other fish stay at sea for three, four, or even five years without spawning, meanwhile growing to a great size, and it is probable that all the exceptionally large fish are maidens entering fresh water for the first time. On the other hand salmon that commence spawning at an early age and spa^vn every year never grow large, for the yearly growth is hardly more than enough to make up for the loss during the sojoiirn in the river. Salmon rarely live more than eight or nine years or spawm more than three or four times. Many (particu- larly, the very large fish) spawn but once, others annually, and others at intervals of two or three years. It follows from this that large salmon are to be fo\md in the sea throughout the year, though fewer of them in summer when the spawning fish are absent about their reproductive duties than in \\'inter when the whole stock, except for the parr and a few spent fish to be mentioned later, is assembled there. Food. — The salmon is purely carnivorous and very voracious, feeding alto- gether on live bait, chiefly on fish and crustaceans. Among the former launce, herring, capelin, smelt, small mackerel, small sculpins, and even flatfish have been described as entering into its diet, with the first three its favorites.*^ Comeau,*" for example, speaks of launce and capelin as having been the chief cUet of thousands of salmon that he opened on the north shore of the Gulf of St. Lawrence. In the Baltic a hook and line fishery is carried on for salmon with herring as bait, and occasionally they have been taken on herring-baited cod trawls off the Maine coast, while herring up to 5 inches in length have been found in salmon stomachs about Eastport. Sand fleas (Gammarus) rank with launce and herring in impor- tance as salmon food in the North and Baltic Seas, while fish entering the Penobscot have been found full of "shrimp" (probably euphausiids) . Salmon are also credited with eating crabs. Most of the large salmon, like the smolts, disappear from the inmiediate neighborhood of the coast in winter, but probably the main body does not go far to sea, for they are regularly caught near land in the Baltic in winter by the hook and fine fishery just mentioned, nor is it unusual for a few salmon to be picked up about Massachusetts Bay at that season, evidence strengthened by the fact that salmon appear about the river mouths so soon after the ice goes out in spring that they can not have come from any great distance. With odd fish entering the rivei « Eichelbaum (Conseil Permanent International pom I'Exploration de la Mer, Rapports et Proces-Verbaiu, Vol. XXI 1916, p. &4) examined the contents of the stomachs of many salmon from the Baltic and North Seas. « Life and Sport on the North Shore, by Napoleon A.. Comeau. 440 pp., iUus., 1909. Quebec. FISHES OF THE GULF OF MAINE 135 mouths in winter (as has often been reported) and with kelts (p. 136) on their way down, salmon are to be found in the large rivers in every month in the year, as has long been known for the Penobscot. There is no reason to suppose that the Gulf of Maine salmon regularly descend to any great depth, winter or summer. On the contrary, the weirs, gill nets, etc., which yield so many in various regions, are all operated in shoal water (the Baltic hook-and-line fishery is carried on at about IJ^ fathoms); but the fact that salmon are caught occasionally on cod trawls in the Gulf is proof that at least some go as deep as 25 fathoms or more, while diet proves they not only feed pelagically, as when pursuing herring, but near if not actually on bottom, where alone they could find Gammarus in abundance. The view now generally held that the whole body of salmon, whether or not destined to breed that season, moves inshore in spring no doubt applies as well to the Gulf as to other seas. Only fish approaching sexual maturity (irrespective of age), and the immature female ''grilse" already mentioned, run far up into the rivers, all others remaining in salt water or at most not rvmning above the head of tide, as has often been remarked. This vernal journey toward the coast takes place long before the spawning season, odd fish even entering the Penobscot in March or earlier, and salmon are to be expected in its lower reaches after the first week of April. Fish apparently coming in from sea are taken off the mouth of Penobscot Bay through May and June and into early July, corresponcUng to the fact that the chief runs in the Penobscot River itself occur in May and June, with a few fish entering even later. Salmon enter the Nova Scotian rivers beginning late in April, and the New Brunswick streams tributary to the Bay of Fundy from May on. In the Shubenacadie grilse are said to run from August until late in autumn. We have not been able to obtain more definite dates for the St. John Eiver. A good deal of discussion has centered about the question as to whether the earliest fish stay in fresh water from then until spawning time, a matter of sLx months, or whether there is more or less movement in and out of the river mouths at the beginning of the season. Probably the latter view is correct, at least for the smaller streams, but it seems safe to say that after the run is well under way in late May or early June no fish return to the sea until autimin. Tagging experiments carried out in Canadian rivers have yielded the very interesting infoi-mation that no matter when a salmon runs upstream one year, it may do so either early or late the next." Wliether or not salmon feed in fresh water has been a muchmootedquestionon which we have no first-hand information to contribute. Certainly, most students and practical anglers believe that salmon feed little in fresh water and not at all as the spawning season approaches, though they may occasionally snap up a minnow or other tempting morsel while still fresh run. The maturing salmon of both sexes lose their silvery sheen in fresh water during the summer months, to take on a dull brownish or reddish hue, while the belly suffuses with some tint of red, large black spots develop, and the male not only becomes variously mottled and spotted mth red ororange,buthisjawselongate,thelowerbecomingsohooked that only the tips come " Fifty-fitth Annual Report of the Fisheries Branch, Department of Marine and Fisheries, Canada. 1921-22 (1922), p. 19. 102274—25} 10 136 BULLETIN OF THE BUREAU OF FISHEEIES together. His body becomes slab-sided, his fins thicken, and his skin is covered witli slime, \mtil altogether he is but a disgusting caricature of the beautiful creature that came in from the sea. Salmon run far upstream — for more than two hundred miles in the larger rivers such as the St. John. SpawTiing, which occupies the fish from 5 to 12 days, takes place in late October and early November in Penobscot and St. John waters; proba- bly at the same time in Nova Scotian streams. Most of the spent fish — now known as "kelts" — return at once to the sea after spawning, this being true of all grilse kelts and of adult fish in small rivers. Probably most of the large salmon that were formerly taken during the winter in Belle Isle Bay, about 30 miles above the mouth of the St. John River at the head of tide, were kelts, for they have been described*^ as improving in condition ^Tith the advance of the season on a bountiful diet of small fish. Some nonbreeding fish seem also to have wintered there,*' how- ever, held by the same attraction. In large rivers, however, some of the kelts linger over ^vinter, taking little food but nevertheless improving somewhat in con- dition, to go back to salt water (if they survive, which many of them do not, for spa\\"ning leaves them verj' thin and exhausted) the follo's^'ing spring, when a few such are caught among the sea-run fish that they pass on their way downstream. None of these late kelts spawn the following autiunn; annual spawners are to be sought among those that run down immediately after spawning and thus have time to recuperate in the sea. Correspondingly it has been found that a far larger pro- portion of the fish in small streams than in large are annual spawners. And here we find one even if not the only reason for the well-known fact that sahnon invariably average smaller in the former than in the latter, for kelts returning to the sea im- mediately after spawning have less opportunity to grow (though they recover condi- tion sufficiently to spawn again the follomng simimer) than such as await the spring to go downstream, and that spend a whole year at sea instead of one winter only between two successive spawnings. This, however, does not account for the fact that it is almost invariably in large rivers that the very large maiden fish, four or more years old, are taken, nor can we subscribe to the oft-advanced explanation that smolts from large rivers wander farther out to sea than do those from small ones and hence are longer in returning thence, for once in the salt water of the open sea, salmon are subject to similar surroundings, irrespective of the size of their parent streams. The distribution of the catch of salmon in the Gulf of Maine yields a glimpse of the movements of the fish there. To begin with, so few are caught near Cape Sable that there can be no general movement around the Cape by the fish that spawn in the rivers of the outer coast of Nova Scotia. The precise locations where salmon are taken in St. Mary Bay (16,400 pounds, or about 1,600 fish, in 1916-17) suggest that the fish follow its southern and not its northern shore on their journeys in and out.^" Statistics of much larger catches in the Bay of Fundy corroborate Hunts- man's (1922a) suggestion that on leaving the rivers they follow the coast (New " Phair. Forest and Stream, Vol. XXX, 1888, p. 291. " Harding, Jas. A. Ibid., Vol. XXXVI, Feb., 1891, p. 68. " A few are also taken near the mouth of the Tusket. FISHES OF THE GULF OF MAINE 137 Brunswick or Nova Scotia, as it happens) for some distance out toward the mouth of the bay and then scatter, to return again in spring by the same route. Thus, to go more into detail, very few are caught on the Nova Scotian side between the entrance to St. Mary Bay and Digby Gut, but fish on their way to the Shubena- cadie River system from the sea yield an abundant harvest as they follow the shores of Annapolis and Kings Counties (the Annapolis River also yields a few fish in its lower course, as well as an odd one occasionally in the Annapolis Basin). Salmon similarly en route to the St. John, the premier Gulf of Maine river, strike the coast about Point Lepreau (about 23 miles west of the St. John), supporting an important fishery from there to the mouth of the river ; but very few St. John River salmon can go up the Bay of Fundy after they leave the river, for hardly any are taken on the New Brunswick shore east of its mouth except for small catches made off the Petit Codiac River near its head, the product of the local run in that stream. A similar trend of salmon out toward the southwest and back along shore in the oppo- site direction seems to obtain on the coast of Maine, for only an occasional fish, probably the product of the Dennys River, is caught east of Mount Desert, while several times as many, which may safely be credited to the Penobscot, are caught along the stretch of coast line from Penobscot Bay to Cape Elizabeth. On the whole, the salmon may be looked upon as rather a local fish in these waters; probably few of them stray very far from the streams to which they resort for spawning, nor is it likely that many of the St. John River or Minas Basin fish ever venture far outside the Bay of Fundy into the open Gxilf. Perhaps a 50-mile radius would encircle the wanderings of the majority except, perhaps, in winter, when we know noticing of their comings or goings. Marked salmon, however, occa- sionally have been known to make long migrations on the other side of the Atlantic, while fish marked in Nova Scotia have been taken in Newfoundland. Proof that salmon may stray equally far in the Gulf of Maine is afforded by the yearly captures near Cape Ann and in Massachusetts Bay of fish that must have come from at least as far as the Penobscot River, if not from the Bay of Fundy, and by the fact that odd salmon, either strays from the Gulf or (less likely) the product of the periodic attempts to restock the Hudson, are caught almost every year about Marthas Vineyard;^' wliile odd salmon, probably of the latter parentage, have been taken in the pounds along the New Jersey coast. Even young fish may travel for considerable distances during their first summer at sea, since "smolts" so small that they must have run down to salt water but a few months previous, and for which no nursery existed nearer than the Penobscot, have been taken in Cape Cod Bay in October (p. 132). No salmon has ever been creditably reported more than about 25 miles from the nearest land in the Gulf of Maine, and therefore the 100- fathom contour incloses practically the entire range of the species there. '^ A question closely bound up with the migrations of the salmon in the sea is: What proportion of them return year after year to spawn in the very rivers in which they were hatched? It has been demonstrated by a variety of evidence, especially " Notably in the spring of 1918 when about 75 (including large fish up to 35 pounds) were taken at Gay Head and in the neighborhood of Woods Hole. '' Smith (1895, p. 99) records one caught 50 miles at sea oti the coast of Delaware. 138 BULLETIN OF THE BUREAU OF FISHERIES by the recapture of tagged fish, that the majority of salmon do so return year after year, a thesis generally known as the "parent stream theory." However, since marked fish have occasionally been retaken in strange rivers, as noted above, and since odd salmon appear yearly in certain streams where none have been hatched for many years — in the Merrimac, for instance — it is equally certain that the "parent stream" theory does not always hold. Probably the truth of the matter is that while the great bulk of the fish never strays far and thus readily returns to the home stream, wanderers that chance to be in the physical state leadmg to maturity when they come inshore in the spring enter any large unpolluted stream they encounter, however far from home. Enemies. — No doubt when they first go to sea the smolts fall prey to any large predaceous fish — they have, indeed, been found in the stomachs of pollock — but after one or two years' sojourn in salt water salmon are so heavy and strong that only fish as large as tuna, swordfish, or the larger sharks can menace them. Their worst enemy is the harbor seal, which is a common denizen of the northeastern coasts of the Gulf of Maine. Fig. 51. — Brook trout {Salvdinus fontinalis) 48. Brook trout {Salvelinus fontinalis Mitchill) Sea trout; Salter Jordan and Evermann, 1895-1900, p. 506. Description. — Although brook trout vary widely in general form in different streams, when taken in salt water they are usually salmonlike in shape — that is, with stout bodies usually about one-fourth as deep as long and tapering gracefully to a small head. Tlae nose of a trout, however, is blunter than that of a salmon, and its head is longer in proportion, the total length of the fish (not coimting the caudal fin) being about four and one-half times that of the head,^^ while its mouth (gaping back of the eye) is relatively larger. The general arrangement of the fins, including the " adipose," parallels that of the salmon, but the ventral fins stand under the mid- dle of the dorsal — that is, they are farther forward in relation to the latter than in its larger relative. All the fins, too, are relatively larger, particularly the ventrals. As a rule the anal has one fewer ray (usually 8) , but the number of dorsal rays .(about 11) is the same. The tail of the sea trout is less forked than that of a young salmon of equal size. •' Some trout are even longer headed. FISHES OF THE GULF OF MAINE 139 Examination of the teeth is the most positive means of distinguishing trout (in European terminology this is a "charr") from sahnon smolt, for the vomerine teeth of the trout are confined to a cUister near the front of the roof of the mouth instead of extending back^vard in a row along its midhne as they do in the salmon. Further- more, the scales of the trout are so tiny as hardly to be visible while those of the salmon are easily seen. Color. — Trout living in salt water are colored very differently from their fresh- water brothers, for they almost wholly lack the yellow and red tints so conspicuous on the latter but are invariably silvery. Sea-run fish are steel blue or bottle green on the back, with cheeks and sides silvery like a salmon and with a wliite belly. The sides above the lateral line are more or less dotted with pale yellow spots, but the dark vermiculate markings so characteristic of the fresh-water brook trout are hardly to be seen on the trunk, though evident as wavy crossbars on the dorsal and on the corners of the caudal fins. Below the level of the lateral line the sides and flanks are strewn with small pale vermilion dots but the ventral fins are plain white, or at most the pink edging so conspicuous in trout caught in fresh water is faint. General range. — Eastern North America, north to Labrador, west to Minnesota, and southward along the Allegheny Mountains to Georgia. Occurrence in the Gulf of Maine. — Trout are plentiful in many of the river systems and smaller streams that empty into the Gulf of Maine. In some of these some of the trout seek salt water after the breeding season, to remain there over the winter. This applies particularly to the brooks that flow through the sands of Cape Cod, several of those on its southern slope being famous for their sea-trout fishing. These, however, lie outside our present province and at present only a couple of small streams on the Massachusetts Bay side of the cape still support a race of trout that rim doMoi to the sea. Sea trout seem to be unknown between Cape Cod and Cape Elizabeth, imless possibly in one or two streams tributary to Ipswich Bay. Without a local knowledge more intimate than we boast we can not say how generally sea trout may now exist in the streams in eastern Maine, but according to Evermann (1905a, p. 105) trout inhabit the tidal portions of many of the brooks that empty into Casco Bay. Huntsman found no definite evidence of trout in salt or brackish water on the New Brunswick side of the Bay of Fundy, but local inquiry has elicited the information that there are fish of this habit in a few streams — notably in Salmon River — on the north and west coasts of Nova Scotia, where in the past many streams formerly held sea trout but have long since been fished out. Anatomically and specifically the "sea trout" is indistinguishable from the ordinary brook trout; ^^ they are simply fish that have the habit of running down to salt water, and even in streams free of access to the sea, cold enough throughout their lengths, and harboring these "salters" (as they are called on Cape Cod), most of the trout never leave fresh water. All who have given special attention to our sea trout are agreed on this. It is still an open question whether the habit is hereditary or whether it is acquired independently by each individual fish. Personally, we incline to the first view, chiefly because sea trout are slow in reestab- lishing themselves in any stream once they are brought to a low ebb by hard fishing. '* There is another species of sea trout in northern Canadian waters, very plentiful along the coast of northern Labrador. 140 BULLETIN OF THE BUREAU OF FISHERIES The trout that follow this habit grow so much more rapidly on the abundant rations the salt estuaries provide than do most of their relatives that remain in the brook that sea fish run up to from 1 to 3 pounds in weight in streams where the fresh- water trout seldom exceed half a pound. On Cape Cod the sea trout go down to salt water in November immediately after spawning, winter there, and begin to run in again in April, all being once more in brackish or fresh water by mid-May. In the Nova Scotian streams tributary to the Bay of Fundy it is said that they do not appear untQ later in spring (we can not vouch for this). While in salt water — at least along Cape Cod — the trout feed chiefly on shrimps, mimimiohogs (Fundulus), and other small fish. Trout never stray far from the stream mouths in the Gulf. So close, indeed, do they hang that we have never head of the capture of a single one outside the tidal creek or estuary into which its home stream empties. Hence trout have no place " in the fish fauna of the open Gulf. THE SMELTS. FAMILY ARGENTINID.S The smelts are small salmons in all essential respects, except that the stomach is simply a sac with few or no pyloric coeca, whereas in their larger relatives of the salmon family there are large numbers of such coeca. However, it is not necessary to look so deeply among the few species proper to the Gulf to tell if a fish be smelt or very young salmon, for the former all have pointed noses and are of slender form, whereas the yoimg of our three salt-water salmons — humpback, Atlantic, and sea trout — are stouter bodied with rounded noses. In most cases, too, the shape of the tail alone would suffice to separate smelt from salmon smolt, for in the latter it is never as deeply forked as in the former, though considerably emarginate instead of square as in the adult salmon. Three smelt fishes occur in the Gulf of Maine — the smelt (very common), capelin (a sporadic visitor from the north), and argentine (rare, but perhaps occur- ring more regularly than actual recorded captures suggest). KEY TO GULF OF MAINE SMELTS 1. Dorsal fin situated far behind pectorals 2 Dorsal fin originates over the tip of pectorals Argentine, p. 147 2. Upper jaw almost as long as lower; teeth large; there is a group of strong fangs on the tongue; pectoral fins have 12 rays or fewer Smelt, p. 143 Lower jaw much longer than upper; teeth so small as hardly to be visible; no fangs on tongue; pectoral fins have 15 to 20 rays Capelin, p. 140 49. Capelin {Mallotus viUosus Miiller) Jordan and Evermann, 1896-1900, p. 520. Description. — The capelin is an even slenderer fish than the smelt, its body being only about one-seventh to one-eighth as deep and about one-twelfth as thick as long, except in the case of females with the abdomen distended with spawn; it -is of nearly imiform depth from gill cover to anal fin, whereas the smelt is usually deepest about its mid length (at least if the fish is fat) , which gives the two species " Trout are occasionally taken about Woods Hole in the nets in winter. FISHES OF THE GULF OF MAINE 141 characteristically different aspects. The head of the capelin is pointed like that of the smelt, the mouth gaping back to below the center of the very large eye with the tipof the lower jaw projecting noticeably beyond the upper. The scales are minute — • much smaller than those of the smelt and more numerous (about 200 per row on the sides of the body) — its teeth so small as to be hardly visible to the naked eye, and the tongue fangs, so characterisitc of the smelt (p. 144), are lacking here. The outline of the adipose fin likewise helps separate capelin from smelt, for in the former it is low and about half as long as the anal, while in the latter it is short and high. The pectoral of the capelin is likewise broader, with 15 or more rays. The capelin exhibits a pronounced sexual dimorphism, males having much the longer pectoral fins, and the base of the anal is elevated on a pronounced hump, whereas it follows the general outline of the belly in the female. In the males, too, a longitudinal row of scales immediately above the lateral line and another along each side of the belly are not only pointed and distinctly larger than the other Fig. 5.5— Adult Fig. 56.— Larva, 23 mOlimetDi'S. After Schmidt C.\PELIN (Maltolus cillosus) scales but become very elongate at spawning time, forming four ridges very evident when the fish is taken in the hand. Color. — The capelin, like the smelt, is transparent olive to bottle green above, but its sides are uniformly silvery below the lateral line and the scales are dotted at the margins with minute dusky specks (in the smelt there is a distinct silvery band on each side) ; the belly is white and at spawning tune back and head darken. Size. — Capelin are seldom more than 63^ to 73^ inches long. General range. — Boreal — Arctic seas, south to the coast of Maine ^' on the Atlantic coast of America. Occurrence in the Qulf of Maine. — The capelin is a sub-Arctic fish that occurs at irregular intervals in the Gulf of Maine, chiefly on its eastern side, as might be expected of a visitor from the north. " According to Jordan and Evermann the capelin finds its southern limit at Cape Cod, but we find no actual records of its occurrence farther south than here mentioned. 142 BULLETIN OF THE BUREAU OF FISHERIES "The capelin (Mallotus villosus)", ^v^ites Doctor Huntsman," "is endemic around Newfoundland and in the Gulf of St. Lawrence. In the latter region, at least, it occurs abundantly only in limited areas, which shift somewhat from year to year. It occurs periodically in similar limited areas farther south. The southeastern corner of Cape Breton is the center of such an area, where large quantities were taken in 1917. Halifax is the center of another area, where, however, it is more rare. In 1916 it was abundant at Sambro,*near Halifax. The next area is in the Bay of Fundy, where they have, exceptionally, been taken in large quanti- ties at long intervals." Apparently a period of this sort occurred about the middle of the past century, for Perley, writing in 1852, reported it from a number of points in the neighborhood of St. John, New Brunswick. It seems then to have disappeared and for many years thereafter was unknown anywhere in the Gulf of Maine, but it reappeared in the Bay of Fundy in May, 1903, when it was common, and a few were again taken off Pas- samaquoddy Bay in that same month of 1915.^^ This was the prelude to a period of local abundance, for capelin were noticed among the herring taken in the weirs of the Passamaquoddy Bay region in October, 1916, becoming so plentiful by the end of November that one catch of 3,000 pounds of fish consisted of 2,000 pounds of capelin and only 1,000 of herring. They were also reported at various localities along the New Brunswick coast at that time. Probably they persisted locally in the Bay of Fundy throughout the winter of 1916-17, for in the following Ma}^ and June large numbers of capelin appeared in Minas Basin. We find no record of capelin within the limits of the bay in 1918, but not only were they taken again in 50 fathoms of water off Passamaquoddy Bay in January, February, and March, 1919, but they appeared with smelts a month later as far west as the Pe- nobscot River, penetrating far inland. Since then none have been seen in the Gulf of Maine. In spite of the fact that the capelin no doubt invades the Gulf from the east, it is unknown along the western shores of Nova Scotia between Cape Sable and the mouth of the Bay of Fundy. Evidently it travels directly north until caught up in the cul-de-sac of the bay on the rare occasions when it rounds the cape, which is in line with other evidence to the effect that, once past Cape Sable, the general entrant track of visitors from the north lies some distance off the Nova Scotian coast. Habits. — Although the capelin is not regularly endemic as far south as the Gulf of Maine it may breed locally in the Bay of Fundy on the rare occasions when it persists there for more than one year. It spawns in salt water (unlike its close relative the smelt, which is anadromous), depositing its eggs on sandy bottom along shore from just below tide mark down to 35 or 40 fathoms, where they stick together in clusters like herring eggs. Many accounts have appeared of the multi- tudes of capelin that gather along northern coasts at this time. *" Quoted from a letter. g It Huntsman (1922a, p. 12) and Kendall (1917, pp. 2S-30, and 1919, pp. 70-71) give details. FISHES OF THE GULF OF MAINE 143 In north European seas capelin spawn from April until July; we have found them spawning along the coast of Labrador in multitudes in July in 1900. Prob- ably April and May would cover such spawning as occurs in the Bay of Fundy. The spawning act has been described variously. According to Lanman (1874, p. 225) each female is accompanied by two males that crowd her between them. The eggs are reddish and about 1 mm. in diameter, while the larvae are about 7 mm. long at hatching. Larvse only 8 mm. in length have entirely absorbed the yolk, the rudiments of the doi'sal and anal fins are visible at 14 mm., while at 23 imii. the adipose fin can be distinguished, the fin rays are formed, and the little capelin is easily recognizable as such. The capelin so seldom appears in the Gulf of Maine that we need state of its habits merely that, unlike the smelt, it is a fish of the high seas frequently encoun- tered far out from land; it never enters fresh water; comes inshore only to spawn and then as a rule moves out again; travels in vast schools at spawning time, Fig. 57.— AdiUt Fig. 5S.— Fry, 26 millimeters SMELT (Osmerus moTdai) when it often strands on the beach in countless multitudes; and that it is the chief bait fish of Arctic seas, preyed upon by whales and every predaceous fish, particularly by cod. Capelin themselves feed chiefly on small crustaceans, partic- ularly copepods, euphausiid shrimps, and amphipods. It is also known to devour its own eggs. DO We can bear witness that it is a delicious little fish on the table. 60. Smelt (Osmerus mordax Mitchill) Salt-water smelt; Icefish Jordan and Evermann, 1896-1900, p. 523. Description. — The smelt is distinguishable from aU other fish common in our waters by its slender form, long pointed head, the presence of a small but evident 144 BULLETIN OF THE BUREAU OF FISHERIES adipose fin standing above the rear part of the anal, and a deeply forked tail. The capelin and argentine alone among the Gulf of Maine fishes share this combi- nation of characters with the smelt, but the latter is distinguishable from the argen- tine at a glance by the location of its dorsal fui above instead of in front of the ven- trals and by its larger mouth, while the large, fanglike teeth of the smelt's tongue, its larger scales (of which there are about 75 rows on the sides, all alike in the two sexes), its shorter adipose fin, its narrower pectoral &a, and the facts that its lower jaw projects but slighlty beyond the upper and that its scales slip off very easUy, obviate all danger of confusing it with the capelin. The body of the smelt is only about one-fifth as deep as long (exclusive of caudal fin), with broadly rounded back but compressed enough to be egg-shaped in cross section. It is deepest about its mid length, tapering thence toward the head as toward the tail (at least in fat fish), whereas the capehn is of nearly uniform depth from giU opening to anal fin (p. 140). Its mouth gapes back of the eye. Printed accoimts of the smelt usually credit it with a peculiar "cucumber" odor, and smelt fishermen often speak of a trace of this, but it is so faint that I have never noticed it though I have caught and handled many.^' Color. — Transparent olive to bottle green above, the sides a paler cast of the same hue but each with a broad longitudinal silvery band. The belly is silvery, while the fins and body are more or less flecked with tiny dusky dots. This color pattern is shared by another slender little fish, the silverside (Menidia, p. 179), but as the latter has two large dorsal fins there is no danger of confusing the smelt with it. Size. — Smelt grow to a maximum length of about 13 or 14 inches. Few larger than a foot long are seen, however, and adults run only about 7 to 9 inches. Accord- ing to size and fatness smelt weigh from 1 to 6 oimces. General range. — East coast of North America from eastern Labrador and the Gulf of St. Lawrence southward regularly to New Jersey, and reported to Vir- ginia, running up into streams and rivers to spawn. Occurrence in the Gulf of Maine. — The smelt is a familiar little fish around the entire coast line of the Gulf of Maine, but varies greatly in abimdance from place to place according to the accessibility of streams suitable for spawning, from which it seldom wanders far alongshore. Smelt are stiU very abundant all around the inner parts of Massachusetts Bay and its tributary harbors, though many of the local streams are barred to them now; thence northward and eastward, too, along the coast of Maine, as illustrated by the following report of commercial catches from the several coast counties of that State in 1919: County Number of pounds County Number of pounds York . 8,000 136,202 74,230 53,500 Knox - .- ._ 71,100 Waldo- - 41,800 Sagadahoo . Hancock- - 91,835 17,850 ' Tbe European smelt (0. eperlanus) smells so strong that it is not held in very high esteem. FISHES OF THE GULF OF MAINE 145 Smelt are also found very generally along both the New Brunswick and the Nova Scotian shores of the Bay of Fundy, but to judge from Canadian fishery statistics they are far more plentiful near the mouth of the bay and on the Nova Scotian side than anywhere on the New Brunswick shore "" or farther up the bay, where only a few hundred pounds are caught annually. Smelts are also plentiful on the west coast of Nova Scotia, facing the open Gulf, as is reflected in a catch of almost 58,000 pounds for Yarmouth County in 1916-17. The adult smelts return to salt water immediately after spawning, to spend the summer either in the estuary into which the stream in which they spa^vn empties or in the sea close by. On the Massachusetts coast north of Cape Cod all the spent fish have left fresh Avater by the middle of May, in some years earlier. On the Maine coast, too, a good proportion of the spent fish are in salt water by the first week in May, for we have seen a bushel of large smelt taken in a weir at Cutler (near the mouth of the Grand Manan Channel) as early as May 4.°' The summer habitat of the smelt varies off different parts of the coast of the Gulf depending on the summer temperature of the water and perhaps on the food supply. In the Massachusetts Bay region and along the southern coast of Maine most of them desert the harbors and estuaries during the warmest season, prob- ably, however, moving out only far enough to find cool water at a few fathoms depth. A few may be found in harbors through the summer, however. Smelt, for instance, are caught in Cohasset Harbor throughout the summer in some years, but not in others; and east of Penobscot Bay, where the surface temperature does not rise as high as it does off Massachusetts, smelt are to be found in the harbors, bays, and river mouths all summer, when they are sometimes taken in numbers in the weirs. The smelts are always confined to a very narrow zone along shore, for none has ever been reported more than a mile or two from land or at more than a few fathoms depth. Though confined to shoal water, the schools (which are mostly composed of individuals of one size and are probably the product of a single hatching) live pelagically, not on the bottom. Food. — The smelt is predaceous. Most authorities describe it as feeding on small crustaceans, which is correct so far as it goes for shrimp (decapods) are probably its favorite food and they are certainly the best smelt bait, but small fish also form an important item in its diet. We have, for example, found smelts taken in the Sheepscot Kiver in May packed full of young herring, and have caught many on small mummichogs (Fundulus), while at Woods Hole cunners, anchovies, and aleA\dves have been identified from smelt stomachs. °^ The Woods Hole diet list also includes shellfish, squid, annelid worms (Nereis), and crabs, but even as greedy a fish as the smelt ceases to feed during its spawning visits to fresh water. Breeding liabits. — The adult smelt all gather in harbors and brackish estuaries early in autumn, smelt fishing with hook and line being in full swing by October, and by the time the first ice forms in December some of them have run as far as the head of tide. The smelt winter between the harbor mouths and the brackish '"The catch in 1916-17 was 115,000 pounds tor Annapolis and Digby Counties, Nova Scotia, wtiile tor Charlotte County, New Brunswick, it was only 7,100 pounds. '• Atkins (1887) gives much interesting information on the smelt in Maine. " Vinal Edwards's notes. 146 BULLETIN OF THE BUEEAU OF FISHERIES water fartiier up — that is, in the same zone as the sea trout (p. 140) — the maturing fish commencing their spawning migration into fresh water as soon as the streams warm to the required degree in spring. Temperature observations by the Massa- chusetts Commission show that the first smelt appear on the spawning beds in Weir River, a stream emptying into Boston Harbor, when the temperature of the water rises to about 40° to 42°. About Massachusetts Bay this may take place as early as the first or as late as the last week in March, depending on the forward- ness of the season and on the particular stream in c[uestion. The chief production of eggs is in temperatures of 50° to 57°, and spawning is completed there by about the 10th or 15th of May, year in and year out. East of Portland smelt seldom commence to run before April, to continue through May as just noted. In the colder streams on the southern shores of the Gulf of St. Lawrence they do not spawn until June, but along the southern New England coast south of Cape Cod, on the other hand, they may commence as early as February. As a rule smelt do not journey far upstream. Many go only a few hundred yards above tidewater, whether the stream be small or large, while some even spaAvn in slightly brackish water, as in certain ponds back of barrier beaches (e. g., Straits Pond, Cohasset, Mass.), but flooding with salt water, which sometimes happens, kills the eggs. The eggs average 1.2 .mm. (0.05 inch) in diameter and sink to the bottom, where they stick in clusters to pebbles, to each other, or to any stick, root, grass, or water weed they chance to touch. According to the Manual of Fish Culture a female weighing as little as 2 ounces will produce between 40,000 and 50,000 eggs; 70,000 eggs have been taken from a fish 73^ inches long."^ The eggs of the closely allied European smelt (Osmerus eperlanus) hatch in 8 to 27 days, according to temperature. Probably the incubation period of the American fish is the same, for smelt eggs are reported as hatching in 13 days at the Palmer (Mass.) hatchery. There is no reason to suppose that the rate of growth of our smelt differs much from that of the European (that is, to a length of 1^ to 2^ inches by the fu'st autumn). The precise season when young smelt go down to the sea in the Gulf of Maine streams is yet to be learned; probably early in summer. Most of the smelt evidently do not spa%vn until they have passed a winter, a summer, and a second winter in salt water. The smelt has proved a favorable fish for artificial hatching and at present large numbers of fry are so produced j'early in Massachusetts, the eggs being taken in Weir River, just mentioned. It has proved possible to reestabhsh smelt by intro- ducing the eggs or fry into streams from wliich it has been extirpated. For example, good smelt fishing was reported in " Poorhouse Brook," Saugus, a tributary of Boston Harbor, three years after the stream was stocked with eggs, and attempts on Long Island have been similarly successful. It is the prevaihng opinion that smelt are now increasing in Massachusetts. To maintain the stock is simply a question of providing spawning grounds of sufficient extent or of making up for the lack of such by artificial propagation. " Eice. Report of the Commissioner of Fisheries of Maryland, 1S78. FISHES OF THE GULF OF MAINE 147 The commercial importance of the smelt, one of our very best fish, is consid- erable, but from a dollars and cents standpoint its sporting value to the coastwise inhabitants of New England is probably greater, for smelt fishermen flock to the harbors and stream mouths throughout the autumn. For instance, as many as 2,326 people have been counted fishing at one time about Houghs Neck in Boston Harbor alone, and this same sort of thing is to be seen up and down the coast. So plentiful are the fish on occasion and so greedily do they bite, especially on the flood tide, that it is usual to number the catch about Massachusetts Bay by the dozens rather than by the individual fish. Slii'imp are the best bait, bloodworms (Nereis) second best, small minnows or clams a poor third, and smelt have also been taken on a small red artificial fly in the Gulf of St. Lawrence. I ^^<^'^^ a. Adult, 6, : Fig. 59. — Argentina (Argentina silni) After Schmidt, c, Larva, 23 millimeters. After Schmidt. d. Fry, 45 millimeter.s. After Schmidt. 51. Argentine {Argentina silus Ascanius) Herring smelt Jordan and Evermann, 1S96-1900, p. 526. Description. — The argentine is a smeltlike little fish with the pointed nose, deeply forked tail, and slender, compressed body characteristic of the family, but it has much larger eyes — a character no doubt associated -with its deep-water home — than either smelt or capelin; its mouth is much smaller, not gaping back even as far as the eye; and its dorsal fin stands wholly in front of the ventrals, instead of above them as in both its near relatives. These characters, together with the presence of an adipose fin above the anal that it shares only with smelt, capelin, the salmon 148 BULLETIN OF THE BL'REAi: OF FISHEEIES tribe, and certain oceanic species such as the pearlsides, among common Gulf of Maine fishes, serve to separate it at a glance from any other species likely to be found within the limits of the Gulf .°^ In the argentine, as in the smelt, the body (about one-fifth as deep as long) tapers toward both head and tail, but its sides are so flat, its back and belly so broadly rounded, that it is nearly rectangular in cross section instead of oval. Its scales, too, are larger than those of the smelt, there being only 60 to 70 rows along the lateral line. The adipose fin is very small and the jaws are toothless, though the palate and tongue are armed with small teeth. Color. — We have never seen a freshly taken adult, and its color is variously described by different authors. All agree, however, that the back is brownish or olivaceous, the sides silvery or with iridescent golden or brassy luster, and the belly white. The adipose fin is said to be yellow."^ Size. — The argentine is a larger fish than the smelt or the capelin, growing to a length of about 18 inches. General range. — North Atlantic, usually in water as deep as 80 to 300 fathoms, chieflj' on the European side where it is known from northern Norway south to the northern part of the North Sea. Occurrence in the Gulf of Maine. — The argentine has seldom been taken in our waters. Eecords have been published for Belfast and for Biddeford Pool, Me., and we collected two specimens on the Grampus — the first, of 49 mm., taken in the clos- ing net at 35 fathoms, S3 miles south of Mount Desert Kock on August 14, 1912, and the second, of 38 mm., one year later, from an open haul from 85 fathoms nearby, which completes the list. Probably it is more common in the deeps of the Gulf than these few captures suggest, for on April 17, 1920, a haul from 200 meters in the southeast corner of the Gulf yielded 43 eggs unmistakably of argentine parentage. Halnts. — Nothing is known of the life of the herring smelt in our Gulf, and little enough is known of it in Scandinavian waters, where it is sometimes caught on deep-set lines baited with herring or mussels, and where, like other deep-sea fishes, it is occasionally swept up to the surface by some upwelUng of the water, to drift there helplessly. It certainly begins spawning as early as April in the Gulf of Maine, and judging from European experience it probably continues aU summer. Although buoyant, the eggs float in the deeper water layers, seldom rising to the surface. They are among the largest of buoyant eggs (3 to 3.5 mm. in diameter), with flat oil globule (0.95 to 1.16 mm.) and vacuolated yolk. Newly hatched larvie averaging about 7.5 mm. long have a large yolk sac, but at a length of 12 mm. this has been absorbed and a line of spots has appeared along the belly. The ruchments of the dorsal and anal fins are visible at 28 mm., and by the time the Httle fish has reached 45 mm. the fin rays are formed, the anus has traveled forward, and the forked out- line of the tail is apparent, but it is not until the larva is about 50 mm. long that the ventral fins appear. ** There is a second argentine in Scandinavian waters, but it is not known on our side of the Atlantic. " For an eicellent account of the argentine see Smitt (Scandinavian Fishes, 1892). FISHES OP THE GULF OF MAINE 149 THE LANTERNFISHES. FAMILY MYCTOPHID^ The most diagnostic external characters of the lanternfishes are their large eyes (situated close to the tip of the blunt snout), wide mouths gaping back beyond the eye, one soft-rayed dorsal fin, a deeply forked tail, and the presence of a series of phosphorescent organs as conspicuous pale spots along the sides. Some of them have an adipose fin on the back behind the dorsal fin, but others lack this. When present, this fin is so small and fragile that it is apt to be destroyed by the rough treatment the fish receive in the tow net, in which they are usually taken. Among GuK of Maine fishes they most nearly resemble the anchovy (p. 124), pearlsides (p. 151), and cyclothone (p. 153) ; but they are readily distinguished from the first of these by the presence of phosphorescent organs and by the fact that the snout does not project beyond the mouth, from the second by their much wider mouths, and from the third by their much larger eyes. They are among the commonest fishes on the high seas, where they live at a considerable depth by day but often rise to the surface at night, and a number of them (especially of the genus Myctophum) have been taken along the continental slope abreast of the Gulf of Maine. Hence one species or another may be expected to stray in over the banks if not into the inner parts of the Gulf. The mmaber and arrangement of the phosphorescent organs offer the readiest means of identification within the family, and should a lanternfish be taken to which the following account does not apply, we refer its captor to Goode and Bean (1896) and to Brauer "" or, better, suggest that it be submitted to the United States Bureau of Fisheries to be named. 52. Lanternfish {Mtho'prora effulgens Goode and Bean) Jordan and Evermann, 1896-1900, p. 566. Description. — This curious little oceanic fish suggests the pearlsides (p. 151) in the presence of an adipose fin and of well-developed phosphorescent organs, but its mouth is much wider (it gapes back of the eye), its eye is even larger and situated close to the end of the nose, and the dorsal profile of its head is very convex. There is a very large and noticeable phosphorescent organ covering the entire tip of the snout, including the margin of the orbit, and extending down over the edge of the upper jaw, to which there is no parallel in the pearlsides or in any fish regularly inhabiting the Gulf of Maine. There are also a number of small phosphorescent spots arranged along the lower sides and on the lower siu'face as shown in the illustra- tion." The arrangement of the fins (all of which are soft) is essentially the same as in the pearlsides except that the dorsal (about 15 rays) and anal (about 16 rays) stand relatively farther forward while the pectoral and adipose fins are proportionately smaller and the caudal fin is more deeply forked. Color. — The color has not been described. Probably it is black or at least of some dark shade, with the phosphorescent organs pale blue or green. » Die Tiefsee-Fische. Wissenschaftliche Ergebnisse der Deutschen Tiefsee-Expedition, 1898-99 (1906), Band XV, Teil I. " The structures along the lateral line shown in the illustration are large scales, not phosphorescent organs. 150 BULLETIN OF THE BUREAU OF FISHERIES FISHES OF THE GULF OF MAINE 151 General range and occurrence in the Gulf of Maine. — This species is undoubtedly oceanic and only a stray within the limits of the Gulf, but one specimen has been found in the stomach of a cod caught on Browns Bank.°^ FAMILY STOMIATID^«» This family includes a large number of soft-rayed oceanic fishes of the mid-depths and of the most diverse appearance, all of them with well developed phosphores- cent organs, very large eyes, large mouths, a lower jaw projecting beyond the upper, and with teeth in both jaws. Some have and others lack the adipose fins, but in all of them the ventrals are inserted far back on the abdomen. They differ from the herrings and salmons in the structure of the skull. Only one species, the pearlsides (p. 151), has been recorded more than once or is to be expected except as a stray in the Gulf of Maine, but two others have been taken there casually. 53. Pearlsides {Maurolicus pennanti Walbaum) Pearlfish; Lanternfish Jordan and Evermann, 1896-1900, p. 526. Description. — The presence of an adipose fin between dorsal and caudal, togetherwith phosphorescent organs, distinguishes the pearlsides (fig. 61, p. 150) from all other fish that occur regularly in the Gulf of Maine. This is a flat-sided, large- headed little fish, its body (about one-fifth as deep as long, excluding caudal fLn) deepest forward of the ventral and dorsal fins, its eye very large, its lower jaw pro- jecting, its mouth large but oblique, and both its jaws armed with minute teeth. The dorsal fin (about 11 or 12 rays) stands above the space between ventrals and the anal, which is longer than the dorsal. The adipose fin (both of Woods Hole '" and of Noi-wegian^' examples) is low and long, much as in the capelin." The caudal fin is broad and slightly forked. The pearlsides has been described as scaleless, but this is not correct, for both Scandinavian and Woods Hole specimens have been found to be clothed with large but extremely thin transparent scales. There is no definite lateral line. The most interesting and diagnostic feature of the pearlsides is the presence of a series of phosphorescent organs or luminous dots situated as follows: First, 12 pau's along the belly between pectoral and ventral fins, followed by 5 or 6 from ventral to anal, and, after a gap, by 24 or 25 between the center of the anal fin and the base of the caudal fin, all these together forming a practically continuous row on each side of the belly from tliroat to tail. Second, a row of larger spots shghtly higher up on each side, 6 from chin to pectoral fin, 9 thence backward to the ventrals. Third, a group of 6 low down on each side of the cheek and throat; likewise a spot in front of the base of the pectoral and 2 on the chin.''' •' Reported by Goode and Bean (1896). ** We follow Boulenger in including the pearlsides, viperfish, and cyclothone in this one family. '• Sumner, Osburn, and Cole, 1913, p. 743. " Smitt, Scandinavian Fishes, 1892, p. 933, pi. 44, fig. 3. " Goode and Bean (1896, p. 96) describe it as "very small," but probably their specimens were battered. " This account is based chiefly on Smitt's description and plate. 152 BULLETIN OF THE BUREAU OF FISHERIES Color. — The pearlsides, according to Smitt, is colored much like a herring, with dark bluish or greenish back and lustrous silvery-white sides and belly. The lumi- nous spots are black rimmed, their centers pale blue in life but turning yellow in alcohol, and there is a narrow black band along the base of the anal fin and from there to the base of the caudal, the latter being barred with a similar black band. Size. — Only 1 to 2}/2 inches long. Oeneral range. — The pearlsides (there are several other species closely allied to it) ranges widely in the open Atlantic, occurring at times in shoals on the coasts of Norway and in British waters. It is particularly common off the coast of Scot- land but has seldom been recorded on the American side of the Atlantic. Occurrence in the Gulf of Maine. — ^The known occurrences of the pearlsides in the Gulf are few. Storer (1867) recordsonefoundaliveonthebeachatNahant,Mass., in December, 1837; another taken from the stomach of a cod at Provincetown; a third picked up alive there in July, 1865 (pictured by Storer on plate 25, figure 5) ; and five others found on the Provincetown beach soon afterward. The pearlsides has not been reported in the southern part of the Gulf since that time, though recorded from Woods Hole on two occasions, 22 having been found dead on the beach in November, 1906. It has been found twice at Grand Manan, however, while in July, 1914, specimens were picked up on the beach at Campobello Island at the mouth of the Bay of Fimdy,'^ and others were taken from the stomach of a pollock caught near by. We suspect that the pearlsides is not as rare in the Gulf of Maine as the paucity of actual records for it there might suggest (in fact, Storer tells us that a Nahant fisherman reported finding them repeatedly in the stomachs of haddock many years ago), but that it keeps out of sight, being an inhabitant of the deeper water layers, as its luminous organs woidd suggest, and comes to the siu-face chiefly at night. It can hardly be plentiful or we would have taken it in our deep tow-net hauls. Habits. — The relatives of the pearlsides are oceanic, living in the mid-depths mostly below 150 fathoms, but the pearlsides itself has so often been foimd in the stomachs of cod and herring (fish that do not descend to any great depth) that there is no reason to regard it as a "deep-sea" stray, nor has it ever been taken far from land so far as we can learn. It probably spawns in early spring, females with large eggs having been taken in Scottish waters in winter. 54. Viperflsh {CTiauliodus sloanei Bloch and Schneider) Jordan and Evermann, 1896-1900, p. 585. Description. — The viperfish has an adipose fin and luminous organs like its relative, the pearlsides, but in general appearance it is very different from it and from all other GuK of Maine fishes. Most obvious of its characteristics is its bull- doglike mouth. The lower jaw is longer than the upper, the latter being armed with four long fangs on each side, while the lower has a series of pointed teeth set far apart, those in front very elongate and all of them so long that they project when the mouth is closed. Furthermore, the snout is so short that the very wide mouth gapes far back of the eye. The body is about seven times as long as deep, " These Fundian records are from Huntsman (1922a, p. 13). FISHES OF THE GULF OF MAINE 153 compressed, deepest close behind the head, and tapering thence evenly to the tail. The very short dorsal fin (6 to 7 rays) stands far forward and its first ray is elongate, half as long as the fish when not broken off, as it usually is. The ventrals are about midway from the snout to the base of the tail fin, variously pictured as either larger or smaller than the dorsal. The small anal is close to the caudal, with the adipose fin over it. The body is clothed with large but very thin scales. There are several longitudinal rows of small luminous organs on the ventral surface, running from throat to tail, and several more such spots on the side of the head, while a great number of very tiny unpigmented dots are scattered over the trunk." Color. — Greenish above; sides with metallic gloss; blackish below. Size. — One foot long. Occurrence in the Guif of Maine. — The only definite Gulf of Maine record is of one specimen found in the stomach of a cod caught on Georges Bank in 1874, but as several have been taken off the continental slope abreast of southern New Eng- land '' in deep water, the viperfish may be expected on the offshore banks as a stray. Fig. 62. — Viperfish (Chauliodus sloarifi) Habits. — Nothing is known of its habits except that it is an inhabitant of the mid-depths of tlie Atlantic Basin and probably never rises closer to the surface than 150 or 200 fathoms except, perhaps, during its larval stages. Its teeth suggest a rapacious habit but there is no actual record of its diet nor of its breeding. 55. Cyclothone {Cyclothone signata Garman) Garman, 1899, p. 246, pi. J, fig. 3. Bescrifition. — Cyclothone, like the pearlsides and viperfish (pp. 151 and 152), is distinguishable from other Gulf of Maine fishes by the possession of phos- phorescent organs. These are arranged as follows : One on the head; 1 close below and in front of the eye; 2 on each gill cover; 9 or 10 between the branchiostegal rays; 2 longitudinal rows on the body, the first containing 13 from throat to ventral fins, 4 from ventrals to anal, and 13 from anal to caudal, whUe the second is a higher row of 7 reaching about as far back as the ventrals. The general aspect of the fish is likewise extremely characteristic, the somewhat compressed body being deepest at the gUl opening with the upper surface of the head " Die Tietsee-Fische, by August Brauer. Wissenschaftliche Ergebnisse der Deutsohen Tiefsee-Expedition, 1398-1899 (1906), Band XV, Teil I, p. 40. " Qoode and Bean (1896, p. 97) list these captures. 154 BULi/ETIN OF THE BUREAU OF FISHEBIES concave in profile, the mouth so large that it gapes back of the eye, the lower jaw projecting, the eye very small, and the gill opening very long. The dorsal fin stands over the anal (the latter is much the longer of the two), both originating close behind the middle of the body. The caudal fin is deeply forked and there is no adipose fin. Color. — -Cyclotkone signata is colorless or a pale gray, except that the blackish, dark silvery lining of the abdominal cavity shows through, that the phosphorescent organs are black rimmed and silver centered, and that there are the following black markings, viz: A Y-shaped mark on the forehead, a series of spots or short trans- verse stripes on the flank, spots between the bases of the dorsal and anal fin rays, one or two transverse streaks across the bases of the caudal rays, and a number of irregular flecks and dots along the back and on the gfll covers. "■ " General range. — This is an oceanic fish, very abundant in temperate latitudes in the Atlantic where it lives pelagic from about 100 down to about 250 fathoms, hundreds having often been taken in a single haul. It is also known from the Pacific. Occurrence in the Gulf of Maine. — Cyclothone appears within our limits only as a stray from the Atlantic Basin, one 23 mm. long taken in a haul from 30 fathoms Fig. 63.— Cyclothone ICeclothone signata). After Brauer on Browns Bank, June 24, 1915, and a second mutilated specimen probably of this species from the Fundy Deep (haul from 90 fathoms), March 22, 1920, being the only definite records. THE LANCETFISHES. FAMILY ALEPISAURID^ The lancetfishes have a very high dorsal fin and a small adipose fin like that of salmon or smelt, a deeply forked caudal, a short anal, large pointed pectorals and ventrals, and a wide mouth with large teeth. Several species are known, all belonging to deep water, only one of which has been taken within the province covered by this report. 56. Lancetfish (Alepisauru-s ferox Gill) Handsawfish Jordan and Evermann, 1896-1900, p. 595. Description. — The combination of high and long dorsal fin with the presence of an adipose fin of itself marks the lancetfish off from all other Gulf of Maine fishes. 77, ;s Per detailed accounts and colored illustrations see Garman (Memoirs, Museum of Comparative Zoology at Har- vard College, Vol. XXIV, 1899, p. 246, plate J, fig. 3. Cambridge), Brauer (Wissenschaftliche Ergebnisse der Deutschen Tiefsee-Eipedition, 1898-1899, Band XV, Teil I, 1906, p. 77, PI. VI, fig. 6), Murray and Hjort (The Depths of the Ocean, 1912, PI. I. London). PISHES OF THE GULF OP MAINE 155 The body is slender, somewhat compressed, deepest at the gill cover and tapering back to a slender caudal peduncle. The snout is long and pointed, the mouth wide, gaping back of the eye, and each jaw has two or three fangs and many smaller teeth. The dorsal fin (41 to 44 rays) originates on the nape and occupies the greater length of the back, is rounded in outline, about twice as high as the fish is deep, and can be depressed in a groove. The adipose fin recalls that of the smelt in form and location. The caudal is very deeply forked and, according to Gunther, its upper lobe is prolonged as a long filament, but if this is a constant feature most specimens so far seen have lost it. The anal fin originates under the last dorsal ray, and is deeply concave in outline. The ventrals are about halfway between the anal and the tip of the snout, while the pectorals are considerably longer than the body is deep and are situated very low down on the sides. There are no scales and the fins are described as exceedingly fragile. Fig. 64. — Lancetfish (Alepi-sauTus feroz) Color. — We find no account of its color in life nor have we seen it freshly taken. Size. — -The collection of the Boston Society of Natural History contains a cast of a specimen about 6 feet long taken off Nova Scotia in August, 1910, and this is probably about the maximum size. General range. — Widely distributed in the deep waters of the Atlantic. Occurrence in the Gulf of Maine. — A specimen brought in by a fisherman from Georges Bank '* about 1878 or 1879 is its only claim to mention here. Others have been taken on the deep slopes of LaHave Bank and of the more easterly fishing banks. This is a deep-sea species, only casual above 200 fathoms. Noth- ing is known of its habits. THE MUMMICHOGS OR KILLIFISHES. FAMILY PCECILID.S The mummichogs are small fishes recognizable by the presence of but one short dorsal fin situated far back and by ventrals situated on the abdomen, com- bined with a small terminal mouth, very thick caudal peduncle, and rounded tail fin. The family is represented by tlu-ee species in the Gulf of Maine, two of Fundulus and one of Cyprinodon, the former slender and the latter deep in out- line, enough difference in body form to distinguish one from the other at a glance. " No information as to this specimen more definite than this is available. 156 BULLETIN OF THE BUBEAU OF FISHEKIES The teeth are likewise different in the two genera, those of Fundulus being sharp- pointed, whereas in Cyprinodon they are wedge-shaped and incisorhke. The two local species of Fundulus are separable by their markings, majalis of all ages being barred with black while the adult heteroclitus is not. 57. Commou mummichog {Fundulus heteroclitus Linnaeus) ■ Killifish; Salt-water minnow; Cobbler; Chub; Mudfish; Mud dabbler Jordan and Evermann, 1896-1900, p. 641. Description. — This is a thick-set, stout-bodied fish, about one-fourth as deep as long, its body thickest just back of the head, whence it tapers to the taU. Both its back and its belly are rounded, but the top of the head between the eyes is flat. The snout, as seen from above, is blunt. The mouth is terminal and so small that it does not gape back to the eye. Perhaps the most striking feature of Fundulus is its very deep caudal peduncle and roimded caudal fin. The fins are of moderate -i^ Fig. 65. — Coiuii.oi. :..Limmichog (Fundulus heteTOditus) size, the dorsal being situated behind the middle of the body above the anal. The pectorals are broad and rounded. Both head and body are covered with large rounded scales. In males in breeding condition the scales on the sides of the head and on the flanks below and behind the dorsal fin develop fingerlike processes, the so-called "contact organs," on their free edges. The mummichog exhibits a striking sexual dimorpliism in the dorsal and anal fins, which are not only larger in the male than in the female, and the anals of a different shape,^° but more muscular (they are used as claspers in the act of spawning) . Color. — Males and females differ in color as well as in the sizes of the fins. Out of breeding season the males are dark greenish or steel blue above, marked on the sides with narrow irregular silvery bars or mottlings made up of a series of dots, and with white and yellow spots. The belly is white, pale yellow, or orange; the dorsal, anal, and caudal fins are dark green or dusky with pale mottlings; the front s" A detailed account of the sexual differences is given by Newman (Biological Bulletin, Marine Biological Laboratory, Woods Hole, Mass., Vol. XII, No. 5, April, 1907, pp. 314-348). FISHES OF THE GULF OF MAINE 157 edges of anals and ventrals are yellow. Sometimes there is a dark-edged, pale- centered eyespot on the rear part of the dorsal fin. At spawning time the pig- mentation of the male is generally intensified, the back and upper sides darkening almost to black, while the yellow of the belly becomes more brilliant and the body generally assumes steel-blue reflections. The females (much paler than the males) are uniform olive to bottle green, darker above, lighter below, without definite markings though their sides often show faint and indefinite crossbars of a deeper tone of the same hue. Their fins are much paler than those of males. Very young fry of both sexes show dark transverse bars on the sides but these are lost with growth. Killifishes, as is well known, vary in shade from very pale to dark, according to the color of their surroundings, and recent experiments *' have proved that their ability to change from light to dark depends on the sense of sight. Size. — The maximum length is 5 to 6 inches, but adult mummichogs are seldom more than 3 or 4 inches long in the Gulf of Maine. Several varieties of this species have been described, but they are so closely allied that it is not necessary to discuss them here. General range. — Coast of North America, from the Gulf of St. Lawrence to Texas. Occxirrence in the Gulf of Maine. — The mummichog is very abundant in suitable locations all around the shores of the Gulf of Maine. However, it seldom if ever ventures into the open sea, for its home is along sheltered sandy beaches, especially where the tide flows over beds of eelgrass or among beds of salt hay (Spartina), among which shoals of "mummies" may often be seen moving in with the flood tide. They also swarm in the tidal creeks that cut up the salt marshes, on the shores of our harbors, and in the brackish water at the mouths of our streams and estuaries, particularly in little muddy pools, creeks, and ditches. So closely, in- deed, do they hug the shore that aline drawn 100 yards out from land would probably inclose 90 per cent of all the mummichogs in the Gulf of Maine. Wliere the shore is bold and rocky, as about the Bay of Fundy, the mummichog is practically re- stricted to brackish water, and often goes up into fresh water. It is not likely that it ever descends to a depth of more than 2 or 3 fathoms in its journeys in and out of the creeks or up and down the shore. It is so resistant to a lack of oxygen, the presence of carbondioxide, and unfavorable surroundings generally, that it can survive in very foul water. Habits. — ^The mummichog winters in a more or less torpid or at least sluggish state on the bottoms of the deep muddy holes or creeks. We have no evidence that it goes to sea during the cold season any more than in summer, and in general it is one of the most stationary of fishes. At ebb tide "mummies" are often trapped in little pools where they remain until the next tide if the water holds, often huddled together in swarms. Should the pool go dry, however, they work their way into the mud for the time being, where we have often found them, and probably they can flop overland for a few yards to some other drain as the striped mummichog does.^^ " Parker and Lanchner. American Journal of Physiology, Vol. LXI, 1922, p. 548. " This habit is described by Mast (Journal of Animal Behavior, vol. 5, No. S, 1915, p. 341-350). 158 BULLETIN OF THE BUREAU OF FISHEKIES Food. — •" Mummies" are omnivorous, feeding on all sorts of edible things, vegetable as well as animal. They have been found full of diatoms, eelgrass and other vegetable matter, foraminifera, shrimps and other small Crustacea, small mollusks, and even with small fish in their stomachs. '^ At spawning time they greedily eat their own or each other's eggs. They soon congregate about any dead fish or other bit of carrion, to prey either upon it or upon the amphipod scavengers that gather on such dainties. Breeding habits. — Spawning probably takes place at the same season in the Gulf of Maine as on the southern coast of New England — that is, June, July, and early August. As sexual activity approaches, the males, now brilliantly tinted, court and pursue the females, rivalry among them being very keen, those most highly colored or most excited usually driving off the others. Sometimes they fight fiercely. At the moment of spawning the male clasps the female with his anal and dorsal fins just back of her anal and dorsal, usually forcing her against some stone or against the bottom, the bodies of both being bent into an S and their tails vibrating rapidly while the eggs and the milt are being extruded.'* Occasionally, however, pairs clasp and spawn free in the water without coming in contact with any object, and sometimes a female is seen to pursue and court a male. They have been seen spawning in a few inches of water, seeking shady spots. The eggs, which are about 2 mm. in diameter, colorless or pale yellowish and siuToimded by a firm capsule, sink and become so sticky on contact with the water that they mass together in clumps or stick fast to sand grains or to anything they chance to rest upon. Incubation occupies from 9 to 18 days, the exact duration probably depending on temperature, this being the factor that governs the rate of development for most fishes. The larva is about 7 to 7.7 mm. long at hatching, its yolk absorbed already, its pectoral and caudal fins fully formed. By the time the little fish has grown to 11 mm. the dorsal and anal fin rays are present in full niunber, and the first trace of the ventrals is to be seen. At 16 mm. the ventrals are apparent, and fry of 20 mm. resemble the adult not only structurally but in form. The mummichog is of some little commercial value as bait, but only locall}'. 58. Striped mummichog (Fundulus majalis Walbaum) Mummichog; Mummy; Ktt.lifish Jordan and Evermann, 1896-1900, p. 637. Description. — This fish closely resembles the more common mummichog in general form, in its sexual dimorphism, in the form of its dorsal and anal fins, and in the development of "contact organs" on the scales of the breeding male; but it is more slender, its snout is more pointed in side view, its body more definitely fusiform, tapering toward both head and tail, and its caudal peduncle is less stout. The most striking point of difference between the two, however, is to be seen in the color pattern, both sexes of Fundulus majalis being definitely barred \vith black s^ Lists of stomach contents are given by Field (1907, p. 29). »< Newman (Biological Bulletin, Marine Biological Laboratory, Woods Hole, Mass., Vol. XII, No. 5, .\pril, 1907, p. 315) gives an interesting account of the courtship and spawning from which the preceding is condensed. FISHES OF TSE GULF OF MAINE 159 at maturity as well as when young. In the male the barring is transverse through- out life, the stripes increasing in number from 7 to 10 in the young to from 14 to 20 in adult fish. In the female, however, the original 7 to 10 transverse bars are trans- formed with growth into two longitudinal stripes on each side, the upper running iminterrupted from gill opening to tail, the lower in two segments, the first from close behind the pectoral to above the ventral, and the second thence backward to just behind the rear edge of the anal fin. Even in the oldest females, however, one or two transverse bars persist on the caudal peduncle. Color. — ^Apart from these black bars the male is dark olive green above with silvery sides, a greenish-yellow belly, and a black spot on the rear part of the dorsal fin; his pectorals and caudal pale yellowish — a decidely paler fish than the other "mummy." At breeding time the males become more brilliant, the back turning almost black, the lower sides and belly changing to orange or golden, and the fins to bright yellow. The female is olive green above and white below. Size. — This is a larger fish than the common mummichog, occasionally grow- ing to a length of 8 inches and frequently to a length of 6 inches. Fig. 66. — Striped mummichog, male {Fundalas majalis) General range. — The coast of the United States from Cape Cod to Florida. Occurrence in the Gulf of Maine. — We have never seen this fish in the Gulf of Maine, though it is very abundant along the southern shores of New England west of Cape Cod. In fact, the only Gulf of Maine records are for the neighborhood of Boston many years ago. Possibly it is more plentiful along the outer sands of Cape Cod than is now realized; north of that, however, it is only a stray, hence we need merely remark that it parallels Fundulus heieroelitus in its confinement to the immediate neighborhood of the land and in its general habits, but that it keeps more strictly to salt water. 59. Sheepshead minnow {Cyprinodon variegatus Lacepede) Jordan and Evermann, 1896-1900, p. 671. Description. — Anatomically, as pointed out above (p. 156), the sheepshead min- now is separated from the mimimichogs by the form of its teeth, which are large, wedge shaped, with tricuspid cutting edges, instead of small and pointed. Further- more, it is so deep bodied (its body nearly half as deep as long, not counting the 102274—25+ 11 160 BUL1.ETIN OF THE BUREAU OF FISHERIES tail fin) that it is not apt to be mistaken. It is a stocky little fish more than half as thick as deep, but flat-sided, with high arched back, small flat-topped head, small terminal mouth hardly gaping back to the forward edge of the eye, and with the thick caudal pedimcle characteristic of its family. Its tail is square (rounded in the mummichogs), and the fact that almost the whole dorsal fin is in front of the anal instead of over it affords another point of difference. The pectorals are large, reaching back past the base of the ventrals, which by contrast are very small. Both body and head are covered with large rounded scales, largest on top of the head and on the cheeks. Young fish are proportionally more slender than old ones. In this species, as in the mummichogs, the dorsal, ventral, and anal fins are higher in the males than in the females. The male, too, is deeper bodied and averages larger. Color. — Out of breeding season both males and females are olive above (males rather darker and greener than females) with pale yellow or yellowish-white belly, Fig. 67.— Sheepshead minnow (^Cyprinodon rariegatus) dusky dorsal, and pale orange pectoral, ventral, and anal fins. The young of both sexes are irregidarly barred with black ti'ansverse stripes, which pereist through life in the female but become obscured in adult males. Females, furthermore, have a black spot on the rear corner of the dorsal fin, which is wanting in males, while the caudal fin of the latter sex is marked by two black cross stripes, one at the base and the other at its margin. In breeding season the male assumes a very brilliant coat, his upper parts turning to steel blue in front of the dorsal fin with a greenish luster behind it, while his belly brightens to a deep salmon, his ventrals and anal change to dusky margined with orange, and his doi-sal shows an orange margin in front. Size. — The largest specimens are about 3 inches long. General range. — Atlantic coast of the United States, Cape Cod to Mexico, in brackish as well as in salt water. FISHES OF THE GULF OF MAINE 161 Occurrence in the Gulf of Maine. — This lish, like so many others, finds its northern limit at Cape Cod and would not deserve mention here at all were it not recorded by Storer from the cape. West and south of Cape Cod, however, as at Woods Hole, it is common enough in very shallow water about the heads of bays, along weedy shores, and in brackish water. We have seined many of them with Fundulus at the head of Buzzards Bay. Habits. — The breeding habits recall those of the "mummy" (p. 158), the males fighting fiercely among themselves and clasping the females just forward of the tail with dorsal and anal fins, while the eggs and milt are extruded. Spawning takes place in shallow water from April to September, the eggs maturing a few at a time, so that any given female spawns at intervals throughout the season. The eggs sink and stick together in clumps by numerous threads. They are 1.2 to 1.4 mm. in diameter, with one large and many minute oil globules. Incubation occupies 5 or 6 days, and even at hatching the larvte (4 mm. long) show alternate light and dark crossbands. At a length of 9 nam. all the fins are formed, and at 12 m.m. the fry show most of the characters diagnostic of the species.^ THE BILLFISHES. FAMILY BELONID.E The most noticeable feature of the billfishes is tliat both jaws are prolonged to form a long slim beak well armed with teeth. Their bodies are very slender, with the anal, dorsal, and ventral fins set far back. There are no finlets, the absence of these being the readiest field mark to separate the billfish from the needle- fish (Scomberesox, p. 1 64) . They are sAvif t-swimming, predaceous fishes, represented by many species, most of them American. Only one has ever been recorded in the Gulf of Maine. 60. Silver gar {Tylosurus marinus Walbaum) Billfish; Salt-water gar; Garfish; Sea pike; and various other local names Jordan and Evermann, 1896-1900, p. 714. Description. — Several genera of fish with very long jaws or "bills" have been recorded in the Gulf of Maine, the silver gar being distinguishable among them by the fact that both its jaws are elongated instead of only the lower as in the half- beaks (p. 163), and that there are no finlets between the dorsal and anal fins and the caudal, whereas in the needlefish (Scomberesox, p. 164) there are five or six small dorsal finlets and as many anal finlets. The long bills and slender bodies give the gars a general aspect so peculiar that they are not likely to be confused with any Gulf of Maine fish other than the two just mentioned. The body of the silver gar is about five and one-half times as long as deep, roimded (not laterally flattened) in cross section, and thicker than deep. Its upper jaw from the eye forward is twice as long as the rest of the head, both its jaws are armed with sharp teeth, and its eye is large. Both its body and the sides of its w An account of courtship and spawning is given by Newman (Biological Bulletin, Marine Biological Laboratory, Woods Hole, Mass., Vol. XII, No. 5, April, 1907, p. 336J and of development by Kuntz (Bulletin, United States Bureau of Fisheries, Vol. XXXIV, 1914 (1916), p. 409). 162 BULLETIN OF THE BUREAU OF FISHERIES head are scaly. The dorsal and anal fins are similar in outline, the anterior rays of both being much longer than those toward the rear. Both fins, too, are situated far back, the dorsal originating slightly behind the forward edge of the anal. The ventrals stand halfway between the eye and the base of the caudal. The latter fin is only slightly emarginate, this fact being the readiest field mark to separate this gar from the only other species of its genus ( Tylosurus acus) so far actually taken near Gulf of Maine limits, for the tail of the latter is deeply forked. There is a distinct ridge or low keel on either side of the caudal peduncle.*" Color. — Greenish, darker above, with silvery sides, dull olive fins, and a dark bar on the gill cover. Scales and bones green. Size. — The silver gar grows to a length of 4 feet. General range. — Maine to Texas; very abimdant on the south Atlantic and Gulf coasts of the United States, often running up fresh rivers above tide water. Fig. 68.— Silver gar ( Ttilosurus marinus) . After Storer Occurrence in the Gulf of Maine. — The silver gar is common enough along the southern shores of New England, e. g., at Woods Hole and in Rhode Island waters, where it is to be found from June to October. Like many other southern fishes, however, it seldom journeys eastward past Cape Cod, the only definite records of it in the Gulf of Maine being of several specimens collected by Dr. William C. Kendall at WoKsneck, Freeport, and Casco Bay, Me., and at Monomoy Island, the southern elbow of Cape Cod. We have not met it within the limits of the Gulf of Maine, nor have we heard even a rumor of its presence there from fishermen, pretty good evidence that it is as rare a straggler as the few records indicate, for large silver gars are not fish to be overlooked. With so little claim to mention here, we need merely note that it is very voracious, feeding on all sorts of smaller fishes, and that it runs inshore, possibly even into river mouths, to spawn. The eggs, described by Ryder, ^' are about 3.6 mm. (one-seventh of an inch) in diameter, and stick together and to any object they may touch, by long threads scattered over their surface. '^ ^ There are many other species of gars in tropical seas, any one of which viitjht stray northward with the Gulf Stream and so to the Gulf of Maine. The silver gar is identifiable among them all by the following combination of characters (no one character alone marks it out among its relatives): Mouth capable of being nearly closed; caudal peduncle with keels; dorsal and anal fins short, the former 15-rayed, the latter 17-rayed; eyes at least one-third as broad as the post orbital part of head is long; body not excessively slender but at least one-flfth to one-sixth as deep as head (including jaws) is long. Jordan and Evermann (1396-1900, p. 709) give a useful key to the species of the family. «' Bulletin, United States Fish Commission, Vol. I, 1881 (1882) p. 283. 8s The closely allied houndfish ( Tylosurus acus Laci'pede) has been taken at Nantucket, but has not been found within the Gulf of Maine. However, since it is not unlikely to appear there as a stray from the south we may point out that it is easily distinguished from the silver gar, which it resembles in general appearance, by its deeply forked instead of only slightly emarginate tail and by the fact that its dorsal and anal fins are much longer, the former with 23, the latter with 21 rays. The following char- acters in combination will serve to Identify it among the several tropical gars: Mouth nearly closable and upper jaw not arched; long dorsal and anal fins; beak at least twice as long as rest of head ; greatest depth of body not more than two-thirds the length of pectoral fin; no lateral stripe. ■FISHES OF THE GULF OF MAINE 163 THE HALFBEAKS. FAMILY HEMIRAMPHID^ The halfbeaks are close allies of the billfish (Belonidse, p. 161), but in the only species of present concern the lower jaw is greatly elongate while the upper jaw is short. They are herbivorous, feeding mainly on green algse, not carnivorous like their relatives. There are many species in warm seas, only one of which is known to reach the Gulf of Maine. 61. Halfbeak (HyporJiamphus roberti Cuvier and Valenciennes) Skipjack Jordan and Evermann, 1896-1900, p. 721. Description. — The most striking feature of the halfbeak, one which is enough of itself to mark it off from every other fish known from the Gulf of Maine, is the fact that while the lower jaw is as long as in the silver gars, the upper is very short.*" This is a slender fish, its body being only one-eighth as deep as long, tapering slightly toward head and tail. As in the silver gar, its dorsal (14 to 16 rays) and anal (15 to 17 rays) fins are situated far back, about equal in length and similar in outline, the former over the latter, without finlets. The ventrals stand about midway Fio. 69. — Halfbeak {Hyporhamphus roberti) between the eye and the base of the caudal. The teeth are small and the scales largest on the upper surface of the head. In young fish the beak is much shorter than it is in adults. Color. — Translucent bottle green with silvery tinge above, each side with a narrow but well-defined silvery band running from pectoral to caudal fin, the sides darkest above and paler below. The tip of the lower jaw is crimson in life, with a short filament, and tlu-ee narrow dark streaks rim along the middle of the back. The anterior parts of dorsal and anal and the tips of caudal fins are dusky. The belly lining is black. Size. — Adults are seldom more than 1 foot long. General range. — Tropical and subtropical on both coasts of America and in the Gulf of Mexico; not uncommon northward to Cape Cod, and straying to the coast of Maine. Abundant ofi' the South Atlantic States. Occurrence in the Gulf of Maine. — In our cool boreal waters the halfbeak is only a rare stray from the south, so far recorded only twice in the Gulf of Maine — that is, from Machias and from Casco Bay, Me. 8" Should a halfbeak be taken in the Gulf of Maine it would be well to consult Jordan and Evermann, for there are several other species that might reach there as strays, either via the Gulf Stream route or from offshore. One, indeed iEulepiorhamphuit vdoi), has been taken at Nantucket. Its lower jaw is even longer and more slender than that of the halfbeak, its body is more compressed, and its pectoral fins are longer. 164 BULLETIN OF THE BUREAU OF FISHEEIES THE NEEDLEFISHES. FAMILY SCOMBERESOCID^ In the needlefishes, as with the billfishes (p. 161), both jaws (of the adult) are elongated to form a slender beak, and the anal, dorsal, and ventral fins are set far back, but the presence of several finlets between the dorsal and anal fins and the caudal in the needlefishes, which the billfishes lack, is a ready field mark for their identification. Furthermore, their teeth are small and weak, and their bodies only moderately elongate. Only four or five species are known in warm seas, one of which is not uncommon in the Gulf of Maine. 62. Needlefish (Scomberesox saunts Walbaum) Billfish; Saury; Skipper Jordan and Evermanu, 1896-1900, p. 725. Description. — The needlefish resembles the silver gar in its slender form and in the fact that both its upper and lower jaws are elongate, but difl'ers from it in the presence of a series of five or six little separate finlets in the spaces between Fig. 70. — Adult needlefish (Scomberesox saurus) Fig. 71.— Needlefish fry (Scomberesox sauriis), about 214 inches long. After Murray and Hjort the dorsal and anal fins and the caudal fin. The body is about nine times as long (not counting caudal fin) as deep, compressed, tapering toward the head and tail, with slender caudal peduncle, and all the fins are small. The dorsal originates slightly behind the origin of the anal, these two fins being similar in outline and standing far back. The ventrals are situated about midway the length of the body. The caudal is deeply forked and symmetrical, much like the tail of a mack- erel. The trunk and a patch on each gill cover are covered with small scales. The lower jaw projects beyond the upper; the teeth are pointed but very small. Color. — Olive green above with a silver band on each side at the level of the eye and about as broad as the latter. There is a dark green spot above the base of the pectoral; the dorsal fin is greenish: the lower parts silvery with golden gloss. In young fry, which live in the surface waters of the open Atlantic, the back is dark blue and the sides silvery. Size.—^p to 18 inches long. Those caught on Cape Cod run a foot and more in length. i PISHES OF THE GULF OF MAINE 165 General range. — Temperate parts of the Atlantic, Pacific, and Indian Oceans, known in the open sea as far north as northern Norway off the European coast, and to northern Nova Scotia"" and the Banks of Newfoundland off the eastern American coast. Occurrence in the Gulf of Maine. — While a straggler to our Gulf from warmer waters offshore or farther south, the needlefish has been taken more often on the northern coasts of New England than have any of its relatives, specifically on Cape Cod, at Provincetown, at several locations in Massachusetts Bay, at Annis- quam a few miles north of Cape Ann, at Old Orchard (Maine), in Casco Bay, at Monhegan Island, and among the islands at the northern entrance to the Bay of Fundy, but we find no record of it along the Nova Scotian shore of the Gulf of Maine. Apparently the inner curve of Cape Cod from Provincetown to Wellfleet is a regular center of abundance for it, as Storer long ago remarked, for schools of billfish are picked up in the traps along that stretch of beach almost every year, the catch rarely amounting to hundreds of barrels, while hosts of them have been known to strand there. Its numbers fluctuate greatly from year to year, however, and it often fails to appear."' As a rule either many or none at all are caught, their appearance being so sporadic that they can not be looked upon as regular summer residents. They are taken any time from mid-June to October or November, the largest catches usually being made late in summer."^ Curiously enough, although skippers are often so plentiful in that particular locality they are so rare farther within Massachusetts Bay that many fishermen from Plymouth to Cape Ann had never heard of them, although others had. Certainly we never saw nor heard even a rumor of the fish in many summers spent at Cohasset, and so far as we have been able to learn it is only a stray in the Gulf of Maine north of Cape Ann. It would not be surprising, however, to encounter a large school anywhere within its limits, for at Woods Hole, where the billfish is ordinarily very rare, it has been taken in large numbers on two occasions (1905 and 190G). Witness, too, its occasional abundance off northern Nova Scotia.°° When it does invade the waters of the Gulf of Maine, it may be expected in multitudes, for it usually travels in vast schools. Day,"^ for example, mentions the capture of 100,000 in a single haul in British waters. Habits. — The skipper is strictly pelagic. So far as known it lives exclusively at the surface, so much so that in English waters, where it is plentiful in summer, few are caught in nets set as deep as a fathom or two. Its hordes are preyed upon by porpoises and all the larger predaceous fishes; cod and pollock, for instance, feed greedily upon them, as do bluefish. When they strand on the beaches, as often happens, it is probably in flight from their enemies. At sea they attempt to escape by leaping, whole companies of them breaking the surface together as has often been described. " Cornish (Contributions to Canadian Biology, 1902-1905 (1907), p. 83) states that large schools can often be seen at Canso skipping over the watar as they flee from the pollock. " Blake (American Naturalist, Vol. IV, Nov., 1870, p. 521) remarked that while years before he saw thousands stranded at Provincetown not one was seen in 1870. It failed in 1921, also, and no doubt in many intervening years. ^2 For recent information on the local abundance of billfish on Cape Cod we are indebted to Capt. L. B. Goodspeed, a fisher- man of long experience and close observation. " The flshes of Great Britain and Ireland, Vol. I, 1880-188!. London. 166 BULLETIN OF THE BUHEAU OF FISHEKIES It is not likely that it ever spawns in the cool waters of the Gulf of Maine, for we have never taken its fry in our tow nets, although they are among the~most numerous of young fish in the open Atlantic between the latitudes of 11° or 12° and 40°N. It certainly spawns in the open sea, probably at the surface. Although its eggs have been described as covered with filaments like those of the silver gars, ^* they are not adhesive like the latter, but pelagic. The most interesting phase in the development of the skipper is that the jaws do not commence to elongate until the fry have attained a length of about 40 mm., and that the lower outstrips the upper at first, so that fry of 100 to 150 mm. look more like halfbeaks (Hemiramphus stage) than like their own parents. Food. — European students tell us that the skipper feeds on the smaller pelagic Crustacea and probably also on small fish, for it is sometimes caught on hook and line. One examined by Doctor Linton at Woods Hole contained chiefly annelids, fragments of fish and vegetable debris, a few copepods, and crustacean larvae. Commercial importance. — The needlefish is not of much commercial importance, being too sporadic in its appearances. However, when large catches are made on Cape Cod they find ready sale to the local Portuguese population. If too many are caught for the local trade to absorb, they are sent to Boston, where they are sold for bait. THE STICKLEBACKS. FAMILY GASTEROSTEID^ Sticklebacks are rather small fish, easily recognizable by the presence of three or more stout free spines on the back in front of the dorsal fin — spines that they can erect or depress at will — and by the fact that each ventral fin is represented by an even larger spine with but one or two rudimentary rays. Bony plates may or may not be developed in the scaleless skin. The GuK of Maine species may be named by the following key: KEY TO GULF OF MAINE STICKLEBACKS 1. Not more than five large dorsal spines 2 Seven or more dorsal spines Nine-spined stickleback, p. 166 2. No bony plates on the upper sides, but there is a bony ridge on either side of the abdomen Four-spined stickleback, p. 171 The upper sides are armed with bony plates, and there is a plate in the midline of the belly, but there are no ridges on the sides of the abdomen 3 3. Many (28 or more) plates on each side Three-spined stickleback, p. 168 Only 5 or 6 plates on a side Two-spined stickleback, p. 171 63. Nine-spined stickleback (Pnngitius pungitius Linnseus) Jordan and Evermann, 1896-1900, p. 745. Description. — The nine-spined stickleback is a slender little fish 5 to 6 times as long (not counting caudal rays) as deep with very slim caudal peduncle, the latter usually with a well-developed keel on either side. Occasionally, however, this keel is very low or wanting. There are no bony plates along the sides of the body, •* Skipper eggs were so described by Haeckel (Archiv fiir Anatomie, Physiologie, und Wissenschaftliche Medecin, herausgegeben von Dr. Johannes Miiller, Jahrgang 1855, p. 23, Taf. V, fig. 15. Berlin) 75 years ago. They were not seen again unti] 1910, when similar eggs, 2.2 mm. in diameter, covered with filaments, were towed in the Atlantic by the Michael Sars (Murray and Hjort. The Depths of the Ocean, 1912). FISHES OF THE GULF OF MAINE 167 these being present, however, along the bases of the anal and dorsal fins and on the caudal keels. There are no true scales. The most diagnostic character is that there are usually 9 spines on the midline of the back (from 7 to 12 have been counted) in a continuous row from just in front of the pectoral to the dorsal fin, leaning alternately to one side or the other and set in a slightly zigzag line. The spines are slightly curved; wider at the base than at the tip; fairly uniform in size, about one- half to one-third as long as the height of dorsal fin ; and each with a small triangular fin membrane at the base. They may be depressed to lie in a shallow mid-dorsal groove. Each ventral fin is represented by a stout curved spine thicker and longer than the dorsal spines. The dorsal and anal fins (the former stands above the latter) are similar in form, tapering from front to rear, the anal preceded by a single stout recurved spine. The tail fin is square-tipped. Size. — Large adults are seldom more than .3, usually 2 to 2i^, inches long. Color. — UsuaUj^ dull olive brown above, the upper sides faintly barred or blotched with darker. The belly is silvery, the pubic and thoracic regions often black. The color varies with the season of the year, with the state of sexual matu- rity, and with the color of the bottom on which it lives, those on dark mud being darker and those on bright sand paler. AU become more briUiant during the breeding season when reddish tints appear under the head, the belly turns greenish, Fig. 72. — Nine-spin'^d stickleback (Pungilius puiKjithts) and black dots develop here and there over the entire body. The male has also been described as assuming a rosy tint beneath. General range. — This is one of the most widely ranging of northern fishes, occurring in both the fresh and salt waters of the northern parts of both hemi- spheres from northern Scandinavia to France on the European coast, from Arctic seas south to New York along the American coast, and westward to Saskatchewan and Alaska. Occurrence in the Gulf of Maine. — This stickleback occurs all around the shores of the Gulf of Maine from Nova Scotia and the Bay of Fundy to Cape Cod, but so far as we can learn it is confined to the brackish creeks in salt marshes, where large numbers may often be taken in company with the mummichogs that swarm in such locations, and where it is to be found throughout the year, and to fresh water. In the Gulf of Maine it seldom or never ventures out into the salt waters of the open sea. About Woods Hole, too, it is distinctly a brackish and fresh-water fish. Hahiis. — Hardly touching our Gulf proper, we need only note that its mode of life and feeding habits are much the same as those of its three-spined relative next to be considered (p. 168), that it is similarly destructive to the spawn and young of other fish, and similarly pugnacious. Probably this stickleback spawns early in 102274—251 12 168 BULLETIN OF THE BUEEAU OF FISHERIES summer'^ on the shores of the Gulf, for in northern Europe its breeding season occurs in June and July. The male often but not always builds a nest attached to grass or weeds in which the female spawns, and he guards nest and eggs until the latter hatch, which occurs in about 12 days, the newly-hatched larvpe being about 6 mm. long. Commercial importance. — This stickleback is of no commercial importance in America, but it is sometimes tried out for oil in north Europe when enough can be caught. 64. Three-spined stickleback {Gasterosteus aculeatus Linnaeus) Two-SPiNED stickleback; Stickleback; Pinfish; Hornpout; Ghoster; Thornfish; Thornback Jordan and Evermann, 1896-1900, p. 747. Description. — The three-spined stickleback is a stouter fish than its nine- spined relative, being about one-fourth as deep as long, and is more compressed, but resembles it in fusiform outline, very slender caudal peduncle, and square tail fin. Its most diagnostic characters are the number of dorsal spines, of which there are three (occasionally four and very exceptionally five) , with the first two usually much the larger, and each with a small triangular fin membrane; the small size of the anal spine (in the nine-spined stickleback this is long and free; in the four-spined long but attached to the fin by the fin membrane) ; and especially the presence of a series of 28 to 33 bony plates on the sides, and of a ventral plate between and behind the ventral fins. The fact that the dorsal fin originates some distance in front of the anal is also diagnostic, while its ventral spines are longer and stouter than those of the nine- spined stickleback. This is one of the most variable of fishes, Smitt "" mentioning no less than 32 "species" or races based on its varieties. Its dorsal spines, for example, may be long or short and vary in number as noted above. Its bonj- plates range from none at all to very well developed ones. Its caudal peduncle may or may not be keeled. Most American authors have recognized an American as con- trasted with a European species at the least, the former supposedly with longer dorsal spines, each of them reaching to the next behind when depressed, and the latter with shorter spines; but inasmuch as the long-spined as well as the short- spined form is known to occur on the other side of the Atlantic, with every possible gradation between the two, and seeing that we ourselves have found both in the Gulf among fish otherwise indistinguishable, we incline to the belief that all the various forms are but environmental races of the one species. And this is well established for the relative strength of the dermal armature, which is M'eak in fresh water and strong in salt water. Color. — This fish is extremely variable in color, a fact hardly mentioned in most American accounts. Fundamentally it is deep graj'ish, olive, greenish-brown, or sometimes blue above, paler and often with silvery reflections on the sides, its belly silvery, and the fins pale, except that the fin membrane is often red. In .breeding season the males are described as turning reddish below from nose to vent " It spawns in April and May at Woods Hole. » Scandinavian fishes, 1892. FISHES OF THE GULF OF MAINE 169 and often up the sides. In females the whole body except the top of the back may then be reddish. At the same time the back turns brownish with transverse bands, and the sides develop brassy reflections. Size. — Maximum length about 4 inches, but seldom more than 3 inches long. It matures at a length of 2 inches. General range. — Coasts and fresh waters of the northern hemisphere, from Labrador to New Jersey on the eastern coast of America and represented on the northwestern coast by a form {Gasterosteus cataphradus) that will probably prove to be identical. Fig. 73— Adult Fig. 74.— Egg Fig. 7.5.— Larva, newly hatched, 4.3 millimeters Fig. 76.— Larva, 6.3 millimeters THREE-SPINED STICKLEBACK {Gasierosims aculeatus) Occurrence in the Gulf of Maine. — This stickleback is very plentiful all around the shores of the Gulf from Nova Scotia to Cape Cod, living indifferently in brackish and in salt water. The ditches and creeks of the tidal marshes, brackish ponds and lagoons, rock pools, and weedy shores in shallow water are its favorite habitats. In such places it may be found practically anywhere, and often in great numbers in company with killifishes and other sticklebacks, for it is the commonest of its tribe in the Gulf, as it is about Woods Hole. It is equally at home in fresh water on the one hand, and in sea water of full salinity on the other. 170 BULLETIN OF THE BUEEATJ OF FISHERIES Like all sticklebacks it is distinctively a shore fish, the great majority of the local stock living their whole lives in estuarine waters. Enough stray out to sea, however, for it to be rather common to pick up a few here and there in the tow net, right out to the center of the Gulf. On such occasions they usually hide in clumps of floating eelgrass (Zostera) or rockweed (Fucus) ; indeed we have learned to expect a stickleback or two whenever we dip up bunches of weeds of any size. These wanderers keep to the surface except, perhaps, in very rough weather. °' Wherever found alongshore it is a permanent all-the-j'ear resident, merely dropping down into slightly deeper water such as the bottoms of the deeper creeks to pass the cold months. In such situations it probably lies in schools in a more or less sluggish condition while the tem.perature is lowest.°^ It is proverbially a pug- nacious fish, using its spines with good eflect as weapons of ofi'ense and defense even on other fishes much larger than itself. Food.— This fish feeds iadiscriminately on the smaller invertebrates, on small fish fry, and on fish eggs, to which it is exceedingly destructive in fresh water. It is not only onmivorous but very voracious, the diet list of specimens examined by Vinal Edwards at Woods Hole including copepods, of which they are often full, isopods, schizapod shrimps, young squid, and some had fed on diatom.s only. Breeding habits. — This stickleback affords the classic instance of nest building and of the care of eggs among fishes, and its nesting has been described so often in popular natural histories that a bare outline will suflice here.'" The spawning time is probably the same in the Gulf of Maine (May and June) as in North European waters,' when the fish assume the nuptial dress described above and the males fight fiercely. It is the male that biulds the nest, selecting for this purpose some sheltered spot in shoal water or in some rock pool. Here he builds a barrel-shaped mass, an inch or so in diameter, of bits of grass, weed, etc., cemented together with mucous threads, which he spins from his kidney's, and weighed down with pebbles. To this nest he escorts one or a succession of females, each of them depositing 100 to 150 eggs in the central cavity. The male then enters the nest to fertilize the ova, which stick in clumps to each other and to the nest. Incubation occupies 6 to 10 days, during which period the male guards the nest, driving away intruders large or small. WTien hatching time approaches, however, he tears down the nest but continues to guard the fry until these can shift for themselves. The young fish are 4.25 to 4.5 mm. long at hatching time. In three or four days the yolk sac is absorbed, when a week old they are almost 8 mm. long, and when 6 weeks old and 14 to 16 mm. long the fry are of adult form with fins and spines fully formed.^ This little fish is of no commercial value in America. In Scandinavia, however, it is sometimes seined in such quantities that it is worth boiling down for oil. •' We have taken this stickJeback on the eastern part of Georges Bank (Mar. U, 1920); over German Bank; in the western basic (station 10307); off Cape Cod; near the Isles of Shoals; off Seguin; and off Matinicus; but in the Bay of Fundy it is known only close to land and off the months of estuaries. 88 Large numbers are sometimes seined in winter in Scandinavian waters. " Smitt (Scandinavian Fishes, 1892) and C. Tate F.egan (The fresh-water fishes of the British Isles, 1911, XXV, 287 pp.. Pis, I-XXXVII) give accounts of the nest building on which the following is based. 1 About Woods Hole it spawns from May until the last week in July. Figures of stages in development of this flsh are given by Kuntz and Eadcliffe (1918, p. 131), A. Agassiz (1S82, p. 2.S8, plate 9) . and by Ehrenbaum (Nordisches Plankton, Band I, 1905-1909, p. 319) FISHES OF THE GULF OP JIAINE 171 65. Two-spined stickleback {Gasterosteus hispinosus Walbaum) ' Jordan and Evermann, 1896-1900 (Gaslerosieus gladiuncvlus), p. 2836. Description. — This stickleback is said to differ from the three-spined stickleback in having a deeper body, fewer rays (10 dorsal and 8 anal), fewer dermal plates (5 or 6 as against 28 to 33), unkeeled caudal peduncle, and a strong cusp at the base of the ventral spine both above and below. Dr. W. C. Kendall informs us that careful examination of large series has convinced him that this is actually a distinct species and not a race of the extremely variable three-spined stickleback, although he saw one specimen apparently intermediate between the two. Color. — In life grass-green, mottled and finely punctated with black on the top of the head and back; sides of head and body golden with dark blotches; breast silvery; ventrals scarlet.'' General range. — Newfoundland to New York. Occurrence in the Gulf of Maine. — Sticldebacks of this type are common in company with the three-spined in Passamaquoddy and St. Mary Bays ^ and in the Bay of Fundy. They may be expected anywhere on the Maine coast, being recorded at Winter Harbor, off Monhegan Island, off Seguin Island, from Casco Bay and its tributaries in both salt and brackish water, and from Kittery. It has also been taken at Swampscott, in Massachusetts Bay, and it is fairly common in summer at Woods Hole. To these coastwise localities we have added tow-net captures off Cape Porpoise, on Platts Bank, in the Western Basin, and on German Bank. Habits. — So far as kno^v^l its mode of life is the same as that of the three-spined species, and sticklebacks of this type have been described as building nests with bits of straw on sandy bottom in New York waters," but so often have the two species or races been confused that nothing more definite can be written of its habits. 66. Four-splned stickleback {Apeltes qtiadracus MitchiU) Bloody stickleback Jordan and Evermann, 1896-1900, p. 752. Description. — The four-spined stickleback lacks dermal plates in its scaleless skin, but a bony ridge on each side of the abdomen makes the fish triangular in cross section,with flat belly and sharp back, and gives it an aspect very different from the other sticklebacks. In side view it is fusiform, tapering to the rather pointed nose and to the slim caudal peduncle. There are three free dorsal spines standing close one behind the other, inchning alternately to one or the other side, and a fourth attached to the dorsal fin by the fin membrane. The anal fin is similarly preceded by an attached spine, and each ventral fin is represented by a stouter curved spine succeeded " This is the Gasterostms biamlealus o( Cuvier and Valenciennes; wheitllandi of Putnam; Qladiunculus of Kendall, but not the Q. bispinosTts of Jordan and Evermann, which is a variety of Q. acuJeatus. < Kendall, 1896, p. 624. • Huntsman. 1922a, p. 13. • See Bean, 1903. 172 BULLETIN OF THE BUREAU OF FISHERIES by about two slender rays. The dorsal fin stands above the anal as in the nine- spined species, but both these fins taper less from front to rear, and the caudal is relatively longer and narrower than in any of our other sticklebacks. Color. — Bro^vnish olive or greenish brown above with dark mottUngs that alternate below the lateral line with the silvery white of the belly. The fin membrane of the ventrals is red. Males are much darker than females. Size. — One and one-half to two and one-half inches long. General range. — An American fish, known along the coast from New Bruns- wick and Nova Scotia to Virginia; at home both in salt and in brackish water and running up into fresh water. Fig. 77.— Adult Fig. 73. — Egg Fig. 79. — Larv.'i, newly hatched, 4.3 millimeters FOUR-SPINED STICKLEBACK (Apeltes quadracus) Occurrence in the ChUf of Maine. — This stickleback is common all around the shores of the Gulf on the Nova Scotian as well as the New England side. We have taken it at Yarmouth, Huntsman (1922a, p. 13) records it from St. Mary Bay and along the New Brunswick shore, well within the Bay of Fundy (Maine has usually been given as its northern limit), and there are many locality records for the coasts of Maine and Massachusetts. It is so much more closely restricted to estuarine situations than is its tliree-spined relative (p. 168) that we have never taken it in our tow nets nor do we find a single record of it in the open sea, but it is a common little fish in the salt marshes of northern New England, where it consorts with other sticklebacks and with mummichogs. Like the three-spined stickleback it often runs up into fresh water, though it is primarily a salt and brackish water fish and is never found far in from the coast. On the south shore of New England FISHES OF THE GULF OF MAINE 173 it is a year-round resident. Probably this is equally true in the Gulf, where it probably gathers in the bottoms of the deeper creeks in winter. So far as known it resembles the three-spined stickleback in its feeding habits (copepods and other small crustaceans being its chief diet) and in its general mode of life. Breeding habits. — In the Woods Hole region this stickleback spawns as early as May and as late as the last week of July, after which spent females are found, but the onset of spawning may be delayed until somewhat later in the cooler waters of the Gulf. The males build a nest of plant fragments, cemented with mucus — a small rudimentary affair, however, compared with that of the three-spined stickle- back — described by Ryder' as less than 1 inch in diameter, conical, with an open- ing at the top. In the manufacture it binds together, by a compound mucous thread which it spins out of a pore near the vent, a few stalks of any water plant, bringing bits of weed or other objects in its mouth from time to time to add to the structure. Finally it picks up the eggs and deposits them in the hollow at the top of the nest. Presumably the male guards nest and eggs during incubation. The latter, which are j'ellow and approximately 1.6 mm. in diameter, sink like those of the other sticklebacks and stick together in clumps. At laboratory temperature (about 70°) incubation occupies six days or thereabouts. Newly hatched larva; are about 4.5 mm. long and similar in appearance to those of the three-spined species but more densely pigmented.' THE TRUMPETFISHES. FAMILY FISTULARIID.S The trumpetfishes are characterized by their slender bodies and tremendously long heads and by the fact that the anterior bones of the skuU are prolonged in a very long tube with the small mouth at its tip. The only other Gulf of Maine species \vith which they could possibly be confused is the pipefish (p. 175). In the latter, however, the tubular snout occupies only about one-eighteenth of the length whereas in trumpetfishes it is nearly one-fourth." Furthermore, the pipefish lacks and the trumpetfish has ventral fins, and the caudal fin of the latter is forked while that of the pipefish is rounded. 07. Trumpetfish {Fistularia tabacaria Linnaeus) CORNETFISH Jordan and Evermann, 1896-1900, p. 757. Description. — The slender body and very long snout of this fish are mentioned above. The body to base of caudal is about 34 times as long as deep and only about two-tliirds as deep as thick. The head occupies almost one-third and the snout about one-fourth of the body length. The bones forming the latter are so loosely united that the snout is very distensible. The mouth is small, situated ' Bulletin, U. S. Bureau of Fisheries, Vol. I, 1382, p. 24. ' The early development is described by Ryder (Bulletin, U. S. Fish Commission, Vol. I, 1882, p. 24) and by Kuntz and Radcliffe (1918, p. 132). 'A specimen of the snipefish ( Macrorhamphosus scolnpaz LinnfEus), a European species with its chief center of abundance in the Mediterranean, was recorded at Provincetown in 1857; otherwise it is not known from the American coast. Should it again stray across the Atlantic it may be recognized by a long tubular snout lilie that of the trumpetfish but a short high body with two dorsal fins, the first consisting of one very stout and serrated spine and four smaller ones. 174 BUL1.ETIN OF THE BUREAU OF FISHERIES somewhat obliquely at the tip of the snout, and the lower jaw projects slightly beyond the upper. The caudal fin is deeply forked and its middle rays are pro- longed in a filament about as long as the snout. Both the dorsal (14 rays) and the anal (13 rays) fins are triangular, higher than long, the former standing exactly above the latter, about two-thirds of the distance back from eye to base of caudal fiij. The ventrals are much smaller — about midway betvreen snout and' tail. The skin is scaleless but is studded with .bony plates or shields. Color. — This fish (we have never seen it alive) is described as reddish broxsTi above, the back and sides with many large, oblong, pale blue spots, the lower surface pale and silvery. Size. — Said to reach a length of 6 feet, but the few specimens that stray north- ward are much smaller. General range. — Tropical; common among the West Indies, rarely wandering northward as far as the Massachusetts Bay region. Occurrence in the Gulf of Maine. — There is only one record of the trumpetfish from the Gulf of Maine — a specimen taken at Rockport, Mass. (north side of Cape Fig. 80. — Trumpetfish {FisiuJaria tabacaria). After Siorer Ann), in September, 1S65, and preserved in the collection of the Essex Institute, where it was examined and identified by Goode and Bean (1879, p. 4). Like other tropical fishes, however, it is not so rare west of Cape Cod, a fevr small ones being taken at Woods Hole almost every year. THE PIPEFISHES. FAMILY SYNGNATHID.a; In the pipefishes the anterior portion of the head takes the form of a long tubular snout with the small mouth situated at its tip, the skin is armed v.ith rings of bony plates, and there is only one dorsal fin (soft rayed) and no ventrals. The snout recalls that of the trumpetfishes (p. 173), but pipefishes differ from them and from most other bony fishes in the structure of their gills, vchich take the form of tufts of small rounded lobes instead of the familiar filaments. In this respect their affinity is with the group of which the sticklebacks are the most familiar exponents. There are many species of pipefishes in warm seas, but only one occurs in the Gulf of Maine. FISHES OF THE GULF OF MAINE 175 68. Pipefish {Sifhostoma fuscum Storer) '" Jordan and Evermann, 1896-1900, p. 770. Description. — This is a ver}- slender little fish, particularly so behind the vent, males being about 35 times as long as deep and females about 30 times. The head is one-eighth to one-ninth the total length (in the trumpetfish it is nearly one- third) ; the snout is tubelike, blunt ended, and vdth the small toothless mouth at its tip. The gill openings are very small. The entire body is covered ^vith an armor of bony plates cormected in rings, of which there are 18 to 20 on the body in front of the vent and 36 to 42 on the tail behind the vent. It is heptagonal in cross- section in front of the vent and hexagonal behind it, a character evident vrlien the fish is in hand. The abdomen of the male is wider just back of the vent than else- where, with two lateral flaps that meet along the midline to form the so-called "marsupial" or brood pouch. The female lacks these. The dorsal fin (36 to 40 rays and 5 or 6 times as long as high) covers 4 or 5 of the bony rings in front of the vent and as many behind it. The caudal fin is rounded, its middle rays the longest. The anal is very small, close behind the vent; the pectorals are of moderate size; there are no ventral fins. ^^i^-'^^^^r^^^^^^is ^^^^^^-^ Fic. 81.— Pipefish (Sipkosioma fuscum) Color. — Greenish or olive above, transversely barred and mottled vdth darker. The lower parts of the gill covers are silvery. The lower sides are sprinkled ^^■ith many tiny white dots, and the angle separating side from abdomen is marked by a longitudinal brown bar. The lower surface of the snout is colorless; thence back to vent pale to golden yellow, -with, the marsupial flaps flesh-colored. Dorsal and pectoral fins are pale, and the caudal is brov,-n.'^ Pipefishes change color accord- ing to the color of their surrounchngs. We have seen them of various shades of olive and bro^vn — even red ones having been described. Size. — Usually 4 to 8 and occasionalh' up to 12 inches long. General range. — Coast of eastern North America, in salt and brackish water, from Halifax, Nova Scotia, to North Carolina. Occurrence in the Gulf of Maine. — Although Cape Ann has often been- set as the northern hmit of the pipefish, in reality it is not uncommon in the Bay of Fundy and has been recorded from many locahties along the coasts of Maine and Massa- w This is the only pipefish that occurs on our northern coasts. For a synopsis ol the various other species of the genus see Jordan and Evermann, 1896-1900, p. 961 n Colors after Storer (185S-1867, p. 412), with which the specimens we have esamined agree in general. 176 buijLetin of the bureau of fisheries chusetts. Its chief home is among eelgrass or seaweeds, not only in salt marshes, harbors, and river mouths, where it often goes up into brackish water, but on more open shores as well. In such locations it is as often caught to-day by boys dipping mummichogs for bait as when Storer wrote of it nearly three-cjuarters of a century ago. The pipefish, like the three-spined stickleback, sometimes strays out to sea on the surface, and while we have never taken it in our tow nets, Kendall (1896, p. 623) has often found it under floating rockweed along the Maine coast. There is no reason to suppose the pipefish is at all migratory, for it is resident in the eel- grass (Zostera) at Woods Hole throughout the year. So far as known pipefishes have few enemies. Perhaps they are protected l^y their noxious smell. They usually propel themselves by the dorsal fin, but when alarmed they can travel swiftly with eel-hke strokes of the tail from side to side. Food. — The pipefish feeds chiefly on minute Crustacea, copepods especially, which are often the sole contents of their stomachs according to Vinal Edwards' experience; also to some extent on fish ova, on very small fish fry, and for that matter no doubt indiscriminately on any small marine animals. Its snout is so distensible that it can swallow larger prey than one might expect. In capturing its prey it has been described as expelling the water from the snout and pharynx by muscular action, depending on the return rush to sweep its victims into its mouth. Breeding habits. — On the southern shores of New England pipefish breed from March to August, and probably through this same period on the shores of the Gulf of Maine. Their breeding habits are so unusual that a whole literature has grown up about them.'^ Since the days of Aristotle it has been known that the pipefish nurses its eggs in the brood pouch (p. 175). It is the male that develops this pouch, the flaps of wluch lie flat against the concave belly out of breeding season, but are swollen and their edges cemented together during sexual activity. At each copu- lation, in wliich the male and female interwine together, the protruding oviduct of the latter is inserted into the opening of the pouch of the former and a dozen or more eggs passed over. A pair of fishes copulate several times in succession — wath in- tervals of rest — until the pouch is filled, the male working the eggs down toward its posterior end by contortions of its body. Fertihzation is supposed to take place during the transference of the eggs from one parent to the other. The eggs become embedded in the lining of the brood pouch, and it has been established for the European pipefish (probably tliis applies equally to our North American species) that the embryo within the egg is nourished by the epithelial lining layer of the pouch, so that the latter functions as a placenta.'^ Incubation occupies about 10 days, according to Gudger, and the young are retained in the brood pouch until they are 8 or 9 mm. long, when the yolk sac has been absorbed. The young pipe- fish are then ready for independent existence, and once they leave the pouch they never return to it, as young sea horses (Hippocampus) are said to do (p. 178). Sev- eral observers agree on this — most recently Miss Marie Poland (now Mrs. C. J. " For a historical survey and a general account of the breeding of the closely allied Siphostoma floridx see Gudger (Proceed- ings," U. 8. National Museum, Vol. XXIX, 1906, pp. 447-500, Pis. V-XI). 1' For detailed (if somewhat divergent) accounts of this interesting phenomenon see Huot (Annale^ des Sciences Naturelles. HuitiJme St^rie, Zoologie, SSrie 8, Tome XIV, 1902, pp. 197-2S8. Paris) and Cohn (Anatomischer Anzeigcr, Centralblatt fur die gesamte wissenscbaftliche Anatomic, Band 24, 1904, pp. 192-199, 3 figs. Jena). FISHES OF THE GULF OF MAINE 177 Fish) , who kept pipefish under observation at the laboratory of the United States Bureau of Fisheries at Woods Hole during the summer of 1922. Pipefish fry kept in aquaria have been found to grow from 10 mm. to 70 mm. in length within about two months after hatching." Probably they mature when about 1 year old. Pipefish may be expected to breed in every favorable locality all around the shores of the Gulf, but there are local differences in this respect, for while St. Mary Bay, Annapolis Basin, and Cobequid Bay, on the Nova Scotian shore of the Bay of Fundy, are breeding centers according to Huntsman, large specimens alone are known about Passamaquoddy Bay on the New Bruns- wick side. No doubt the estuarine waters from the Massachusetts Bay region to Penobscot Bay are favorable nurseries. Commercial importance. — The pipefish is of no commercial importance. It is not even good for bait. THE SEA HORSES. FAMILY HIPPOCAMPIDiE 69. Sea horse {Hippocampus hudsonius DeKay) Jordan and Evermann, 1S96-1900, p. 777. Description. — With its laterally compressed body, its deep convex belly, its curved neck and curious horselike head carried at right angles to the general axis -^.' . -^'^^^^^^^^m ^M$.^ Fig. 82. — Sea horso (Hippocampus hudsonius) of the body, the sea horse grotesquely resembles the "knight" in an ordinary set of wooden chessmen. The head is surmounted by a pentagonal star-shaped coronet, and the snout is tubular with the small oblique mouth at its tip, like that of its relative the pipefish. There is a sharp spine on each side above and one behind the eye, a third over the gill cover, and a fourth on the side of the throat, which some- times terminate in cirri, besides a blunt horn between the nostrils. Neck, body, and tail are covered with rings of bony plates, 12 rings on the trunk, 32 to 35 on the tail, and each body ring is armed with four blunt spines. The dorsal fin (about 19 rays) originates about midway of the length of the fish, opposite the vent, and runs backward over three and one-half rings — that is, to within half a ring of the commencement of the tail. The very small anal stands opposite the posterior » Tracy, 1910, p. 93. 178 BULLETIN OF THE BUREAU 01" FISHERIES part of the dorssil. The pectorals are of moderate size, broad based and round tipped; there are no ventrals and no caudal fin. The body tapers suddenly behind the anal fin to a long tail, which is four-cornered in cross section, curled inward, and strongly prehensile. In the male the lower siu-face of the fore part of the tail bears the brood pouch opening by a slit in front. Color. — Light brown or dusky to ashen gray or yellow, variously mottled and blotched with paler and darker — sometimes spangled with silver dots, sometimes plain colored. European sea horses change color according to their surroundings, tints of red, yellow, brown, and white all being wtliin their capabilities, and it is probable that the American species is equally adaptable. Size. — Adults are usually 3 to 6 inches long, one of 734 inches being the largest on record.^^ General range. — Atlantic coast of North America, occurring regularly from South Carolina to Cape Cod, and to Nova Scotia as a stray. Occurrence in the Gulf of Maine. — Although an occasional specimen has been picked up on Georges Bank and as far east as Nova Scotia, the sea horse is not common much beyond New York. Only a few are found each year about Woods Hole, chiefly in July, August, and September, and they so rarely stray past Cape Cod that we have only one definite (Provincetown) and one dubious (Massachusetts Bay) record of its capture in the Gulf of Maine, dead or alive. Sea horses dwell chiefly among eelgrass and seaweed,'^ where they cling with their prehensile tails, monkeylike, to some stalk. They usually swim in a vertical position by undulations of the dorsal fin, not with the tail, the trunk being too stiff for much lateral motion. Food. — Sea horses feed on minute Crustacea and on various larvae — in fact on any animal small enough — sucking in the prey as does the pipefish (p. 176). Habits. — These fish breed in summer *' and the breeding habits resemble those of the pipefish, the male nursing the eggs in his brood pouch where they are deposited a few at a time by the female in repeated copulations. At hatching the young, of which there may be as many as 150, are about 10 to 12 mm. long. When the yolk sac is absorbed the father squeezes them out of the brood sac. According to some students they swim out and in at will, but this calls for verifica- tion. Within a few days after they are set free they already resemble the adult in general appearance. Commercial importance. — The sea horse is of no commercial value but is an object of constant interest to visitors to marine aquaria. THE SILVERSIDES. FAMILY ATHERINIDiE These are small fishes, smeltlike in appearance but with a spiny as well as a soft dorsal fin and with no adipose fin. Two species are known from the Gulf of Maine. KEY TO GULF OF MAINE SILVERSIDES 1. About 24 rays in the an."J fin Common silverside, p. 179 Only 15 or 16 rays in the anal fin Waxen silverside, p. 181 » Bulletin, New York Zoological Society, Vol. XVI, No. 66, Mar., 1913, p. 972. " Gill (Proceedings, U. S. National Museum, Vol. XXVIII, 1905, pp. 805-814) has given an excellent account of the habits and life history of the sea horse. " Ryder (Bulletin, U. S. Fish Commission, Vol. I, 1881 (1882), pp. 191-199) describes its development. FISHES OF THE GULF OF MAINE 179 70. Silverside ( Menidia notata Mitchill) YouNQ smelt; Green smelt; Sand smelt; Whitebait; Capelin; Sperling; Shiner Jordan and Evermann, 1896-1900, pp. 800, 2840. Description. ^This silvery little fish is often confused with the young smelt, but it does not require very close examination to tell them apart for the adipose fin characteristic of the smelt is lacking in the silverside, while the latter has a spinous as well as a soft dorsal fin instead of one dorsal only as in the smelt, this last character distinguisliing it equally from young herrings. The silverside is a slender fish, about one-sixth as deep as long, not counting caudal fin; thin-bodied but with rounded, not sharp-edged, belly; with short head, large eye, and small mouth .^,^«S^ -^*£i^ riG. 83.— -V Jolt Fig. 86.— Fry, 13 millimeters SILVERSIDE (.Uenidia notata) set very obliquely. Both head and body are clothed with large scales. The first dorsal fin (5 spines) is smaller than the second and originates about midway between the tip of the snout and the base of the caudal fin; the second dorsal has 8 to 10 soft rays. The vent is under the middle of the first dorsal, and the anal fin origin- ates under its last spine. The anal (of 24 rays, the first stiff and the others soft) is falcate in outline. The caudal peduncle is slender and the tail forked.''* >' The common silverside is represented on the coasts of the eastern United States by two races, a southern and a northern, not, however, very distinct and connected by such various intergradations that they hardly deserve the two names with which they are usually dignified. The southern form has fewer scales than the northern, only 4 instead of 5 spines in the first dorsal, and is rather a stouter-bodied fish. Kendall has eiven an account of the genus in Report .Commissioner of Fish and Fisheries, 1901 (1902), p. 241. 180 BULLETIN OF THE BUREAU OF FISHEKIES Color. — Translucent bottle green above, with top of head, nose, and chin dusky. The upper sides are thickly specked with dark brown, and there is a silver band outlined above by a narrow black streak, running along each side from close behind the pectoral fin to the base of the caudal. The belly is white. Size. — The silverside grows to a length of 6 inches or more, adults usually being 4 or 5 inches long. General range. — The northern variety of the common silverside is known from Halifax to the Capes of Delaware, south of which it gives place to intergrades or to the southern form, and the latter in its turn has been detected as far north as Woods Hole, but never east of Cape Cod. Occurrence in the Gulf of Maine. — The silverside is to be found all aroimd the shores of the Gulf from Nova Scotia to Cape Cod, always, however, closely confined to the coast line and as a rule within a few yards of the tide line. There is no reason to suppose that this fish ever ventures out to sea or descends deeper than a fathom or two. Many summers spent on the coast leave us with the impression that, generally speaking, the silverside is neither as omnipresent nor as abundant in the Gulf as it is south of Cape Cod. However, great schools of them are often to be seen along the sandy beaches, particularly in Cape Cod Bay and here and there on the Maine coast. Bushels have been caught in a single haul of the seine in Casco Bay and very likely could be elsewhere, but silversides are seldom seen along the stretches of rocky coast exposed to the open sea, which make up a large part of the shore line of the Gulf of Maine. Silversides are extremely gregarious, congregating in schools usually made up of even-sized individuals. They frequent sandy or gravelly shores chieflj', and at high tide are often seen among the sedge grass (Spartina), where it grows sparsely between tide marks, particularly about the inner bays and in river mouths where they follow the tide up and down the beach within a few yards of the water's edge. They also run up into brackish water. The Bay of Fundy affords a good example of the influence the character of the shore line plays in determining the distribution of silversides, for according to Huntsman they are chiefly restricted to brackish water about St. Andrews but are more generally distributed on the New Brunswick shore further up the bay and on the Nova Scotian side as a whole. Silversides are probably resident throughout the year wherever found. Such, at least, is the case in southern New England. Food. — Silversides are omnivorous, feeding chiefly on copepods, mysids, small shrimps, amphipods, fish eggs (including their own!), j^oung squid, annelids, Clado- cera, molluscan larvae, and young prawns. Insects that fall into the water have also been found in their stomachs, as have algte and diatoms mixed with sand and mud. Breeding hahits.^^ — Silversides spawn in Ma}^, June, and early July on the south- ern New England coast. Spawning may commence a little later in the Gulf of Maine, corresponding to lower temperature. The fish then gather in shoals to deposit their eggs on sandy bottom, often among the sedge grass or even above low- " Kuntz and RadcliSe (1918, p. 127) describe its development, and Hildebrand (Bulletin, U. S. Bureau of Fisheries, Vol. XXXVin, 1921-22 (1923) that of the southern race. PISHES OF THE GULF OF